THE GENERATION AND EXPERIMENTAL STUDY OF MICROSCALE ...

THE GENERATION AND EXPERIMENTAL STUDY

OF MICROSCALE DROPLETS IN

DROP-ON-DEMAND INKJET PRINTING

LI ERQIANG

(B.Eng., Xi’an Jiaotong University)

A THESIS SUBMITTED

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF MECHANICAL ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2010

Acknowledgements

i

Acknowledgements

First I would like to express my deepest appreciation to my advisor Professor

Jerry Fuh Ying Hsi for his guidance and supervision throughout this project.

This thesis would never been written without his continuous support and

encouragement. He is very helpful, generous and is very considerate of and

patient with his students. Becoming his student is my great honor.

I most sincerely thank my co-advisor Professor Wong Yoke San, for his

constructive guidance and valuable time on my research. He is very kind,

helpful, considerate, enthusiastic and productive. Furthermore, his hands-on

approaches for research will have a lasting impact on my career in the future.

I would like to express my deepest appreciation to my co-advisor Professor

Sigurdur Tryggvi Thoroddsen, for his continuous support, endless encourage,

constructive guidance and supervision throughout this project. I have learned

from him not only knowledge but also rigorous attitude towards scientific

research.

I am very grateful to Associate Professor Loh Han Tong for his concern and

suggestions in project related issues.

My sincere thanks go to Dr. Zhou Jinxin for his support and enthusiastic

encouragement. During nearly the whole process of my research, he gave me a

lot of advice and help. My sincere gratitude should also go to Dr. Sun Jie, Dr.

Wang Furong, Dr. Feng Wei, Miss. Xu Qian, Miss. Wu Yaqun, Mr. Thian

Chen Hai Stanley, Mr. Zhang Fenghua, Mr. Wang Shouhua, Mr. Ng Jinh Hao

and Mr. Yang Lei for their assistance and knowledge in carrying out the

project.

I had the privilege of working with exceptional students from the department,

including Chang Lei, Li Jinlan, Tan Wei Qiang Emil, Wu Yong Hao Benjamin,

Tan Eng Khoon, Ng Lai Xing, Shareen Chan and Lim Wei Ren Farand. They

have all worked together with me and given me great help in the development

of my research project. They are also my friends and made my graduate study

in Singapore colorful and memorable.

My sincere gratitude should also go to the members of the Fluid Mechanics

Lab, Advanced Manufacturing Lab (AML), Workshop 2 (WS2), Impact

Mechanics Lab, Tissue Engineering Lab, Cellular and Molecular

Bioengineering Lab, and the various Laboratories and Workshops of IMRE

and NUS and their technical staff for their support and technical expertise in

overcoming the many difficulties encountered during the course of the project.

Lastly, but most important, I would like to thank my grandparent, my parents,

my brother, and my girl friend Li Xinxiu (all I can say is that I have the best

girl I could ever hope to have), for their unconditional love and support. They

always believe in me and have done all they can to support my choices.

Table of Contents

ii

Table of Contents

Acknowledgements .............................................................................................. i

Table of Contents ................................................................................................ ii

Summary ............................................................................................................ vi

List of Tables ...................................................................................................... x

List of Figures .................................................................................................... xi

List of Symbols ................................................................................................. xx

1. INTRODUCTION..................................................................................... 1

1.1 Background .............................................................................................. 1

1.2 Challenges ................................................................................................ 5

1.3 Objectives ................................................................................................. 7

1.4 Organization ............................................................................................. 8

2. LITERATURE REVIEW ...................................................................... 10

2.1 Introduction to Inkjet Printing ................................................................ 10

2.1.1 Classification of Inkjet Printing Techniques ................................... 10

2.1.1.1 Continuous Inkjet Printing ........................................................ 11

2.1.1.2 Drop-on-Demand Inkjet Printing .............................................. 14

2.1.2 Advantages and Disadvantages of Inkjet Printing ........................... 21

2.1.3 Printing System Evaluation ............................................................. 23

2.1.3.1 Print Resolution ........................................................................ 23

2.1.3.2 Jetting Frequency ...................................................................... 24

2.1.3.3 Drop Positioning Error .............................................................. 25

2.1.3.4 Nozzle Hydrophobicity Treatment ........................................... 26

2.1.3.5 Inkjet-Printed Droplet Feature after Drying ............................. 27

Table of Contents

iii

2.1.3.6 Inkjet-Printed Line Morphology ............................................... 30

2.2 Squeeze Mode Piezo-Driven Printhead ................................................. 32

2.2.1 Theory of Droplet Formation ........................................................... 32

2.2.1.1 Principle of Squeeze Mode Piezo-Driven Printhead ................ 32

2.2.1.2 Droplet Generation Conditions ................................................. 35

2.2.1.3 Droplet Velocity and Droplet Size ............................................ 39

2.2.1.4 Satellite Droplet ........................................................................ 41

2.2.2 Printhead Fabrication ....................................................................... 45

2.2.2.1 The Overall Printhead Structure ............................................... 45

2.2.2.2 Ejection Nozzle Requirements .................................................. 46

2.2.2.3 Ejection Nozzle Fabrication Methods ....................................... 47

2.3 Creation of Ultra-Small Droplets ........................................................... 52

2.3.1 Needs for Generation of Ultra-Small Droplets ................................ 52

2.3.2 Methods for Printing Ultra-Small Droplets ..................................... 55

2.3.2.1 Reducing Nozzle Size ............................................................... 55

2.3.2.2 Controlling of Waveform .......................................................... 55

2.3.2.3 Electrohydrodynamic Jetting .................................................... 58

2.4 Organ Printing - Science Rather Than Fiction ....................................... 62

2.4.1 How to Realize ................................................................................. 63

2.4.2 Challenges and Requirements .......................................................... 69

3. NOVEL PRINTHEAD DESIGN ........................................................... 72

3.1 Introduction ............................................................................................ 72

3.2 Printhead Fabrication ............................................................................. 74

3.2.1 Printhead Chamber .......................................................................... 75

3.2.2 Interchangeable Nozzle Design ....................................................... 78

Table of Contents

iv

3.3 Experimental Testing of the New Printhead .......................................... 83

3.3.1 Experimental Setup .......................................................................... 83

3.3.2 Experimental Conditions ................................................................. 86

3.3.3 Testing Liquids ................................................................................ 87

3.4 Experimental Results ............................................................................. 89

3.4.1 Comparison of PET/PTFE-Based and Glass-Based Printhead ........ 89

3.4.2 Effect of Pulse Width ....................................................................... 91

3.4.3 Effects of Voltage Pulse Amplitude ................................................ 94

3.4.4 Nozzle Size ...................................................................................... 96

3.4.5 Repeatability .................................................................................... 97

3.4.6 Maximum Jetting Frequency ........................................................... 98

3.4.7 Jetting of Non-Newtonian Liquid .................................................. 101

3.5 Conclusions .......................................................................................... 104

4. FORMING A FINE JET IN INKJET PRINTING ............................ 106

4.1 Introduction .......................................................................................... 106

4.2 Experimental Setup .............................................................................. 108

4.3 Experimental Results ........................................................................... 108

4.3.1 Jet I ................................................................................................. 108

4.3.2 Type II Jetting from Entrained Bubble .......................................... 111

4.3.3 More on Surfaces Collapse Jets ..................................................... 124

4.3.4 Viscosity Effects on Jet Velocity ................................................... 126

4.3.5 Relationship between Jet Velocity and Jet Diameter ..................... 128

4.4 Conclusions .......................................................................................... 130

5. CELL PRINTING ................................................................................. 132

5.1 Introduction .......................................................................................... 132

Table of Contents

v

5.2 Material Preparation and Experimental Procedure .............................. 135

5.2.1 Preparation of Cells, Alginate and Collagen ................................. 135

5.2.2 Printing Experimental Setup .......................................................... 136

5.2.3 Survivability Tests ......................................................................... 139

5.3 Results and Discussion ......................................................................... 140

5.3.1 Cell Survivability Study ................................................................. 140

5.3.1.1 Cell Printing ............................................................................ 140

5.3.1.2 Cell Survivability: Effects of the Mean Shear Rate ................ 142

5.3.2 The Number of Cells in Each Droplet ........................................... 146

5.3.3 The Location of Cells inside Each Droplet .................................... 151

5.3.4 Printing Patterns ............................................................................. 153

5.5 Conclusions .......................................................................................... 156

6. RECOMMENDATIONS FOR FUTURE WORK ............................. 158

6.1 Printhead Design .................................................................................. 158

6.2 Reducing Droplet Size ......................................................................... 159

6.3 Cell Printing ......................................................................................... 159

Bibliography ................................................................................................... 161

Publications ..................................................................................................... 176

Summary

vi

Summary

For environmental conservation and the realization of a sustainable society, it

is necessary that industrial manufacturing processes undergo a transformation

with reduction of environmental impact. From this viewpoint, additive

manufacturing technologies have attracted considerable attention because they

have the potential to greatly reduce ecological footprints as well as the energy

consumed in manufacturing. Inkjet printing is one of the most successful

additive manufacturing technologies. It develops at a rapid pace and has been

expanded from conventional graphic printing to various new applications,

such as organ printing, displays, integrated circuits (ICs), optical devices,

MEMS and drug delivery. Accordingly, the dispensed liquids have been

expanded from the conventional pigmented ink (or standard dye-based ink) to

polymers, gels, cell ink or other materials which often have higher viscosities

or even contain large particles or cells. Consequently, the traditional inkjet

printer designed for graphic printing is unable to fulfill the new challenges,

one of which is to dispense fluids of very high viscosities. For most of the

commercial inkjet printheads, only liquids with viscosities lower than 20 cps

can be consistently dispensed. Fluids with even higher viscosities have to be

diluted before printing or warmed up during the printing, which will adversely

affect the properties of the liquids. Another challenge is raised by nozzle

clogging. Fluids containing particles, or cells, can easily block the nozzle

orifice, resulting in time-consuming nozzle cleaning or even damage of the

entire conventional printhead. To solve the problem, the easiest way is to use a

nozzle with a bigger orifice, as bigger orifices are less likely to clog. However,

Summary

vii

this is often not desirable in inkjet printing as bigger nozzles result in bigger

droplets and lower printing resolution. The poor printability and nozzle

clogging may result in unreliable or failed dispensing when using the

traditional inkjet printhead design for complex liquids.

In this research, a PET/PTFE-based piezoelectric DOD inkjet printhead with

an interchangeable nozzle design was proposed and fabricated by the authors.

The printhead chamber is made of PET or Teflon tube, which is much softer

than the commonly used glass tube. The ejecting capacity of this novel

printhead was compared with commercial printheads, and found to have

superior performance and versatility. Our printhead succeeded in dispensing

aqueous glycerin solutions with viscosity as high as 100 cps, while the

corresponding commercial printheads could only dispense liquids with

viscosities lower than 20 cps. PTFE-based printhead provides excellent anti-

corrosive property when strongly corrosive inks are involved. The

interchangeable nozzle design largely alleviates the difficulty in cleaning of

clogged nozzles and greatly reduces the occurrence of printhead damage. The

effects of operating parameters, including voltage pulse amplitude, pulse

width and jetting frequency, on droplet size and droplet velocity were

characterized. The new printhead shows excellent repeatability.

The formation of fine jets during the piezoelectric drop-on-demand inkjet

printing was investigated using ultra-high-speed video imaging. The speed of

the jet could exceed 90 m/s, which was much higher than the general droplet

velocity during inkjet printing. The diameters of the thinnest jets were of the

Summary

viii

order of a few microns. The generation of such fine jets was studied over a

wide range of viscosities, using 7 different concentrations of water-glycerin

solutions. This jetting was associated with the collapse of an airpocket which

was sucked into the nozzle during the printing. This occurred for longer

expansion times for the piezo-element. Two types of jet were identified during

the printing. The relationships between the speed of the fine-jet and other

parameters like the diameter of the jet and the physical properties of the liquid,

were also characterized. The study provides a possible way to improve inkjet

printing resolution without reducing nozzle diameter.

The in-house-developed printhead was also used for cell printing. The study

has demonstrated that piezoelectric DOD inkjet printing is able to successfully

deliver L929 rat fibroblast cells through nozzles as small as 36 µm. There was

no significant cell death when dispensing the cells through the 81 µm and the

119 µm nozzle, with the mean survival rates only reducing from 98% to 85%.

This is in good agreement with the existing study, in which a commercial

printer was used to print human fibroblast cells. When the orifice was reduced

to 36 µm, the corresponding cell survival rates fell from 95% to 76% when the

excitation pulse amplitude increased from 60 V to 130 V. These results

indicate that the droplet ejection out of the nozzle has exerted large shear

stresses on the cells and possibly disrupted the cell membrane and killed about

20% of the cells. Mean shear rate was estimated by combining the effects of

droplet velocity and orifice diameter and was correlated with the cell survival

rate. A large range of mean shear rates from 1.3×104 s

-1 to 9.2×10

5 s

-1 were

generated and cell survival rates were found to be strongly affected by the

Summary

ix

higher mean shear rates, especially when the shear rate exceeds 5×105 s

-1. The

distribution of the number of cells within each droplet was also investigated.

This was done to find out the minimal cell concentration in the medium, which

is required to avoid the appearance of empty droplets, since droplets

containing no cells may be detrimental to pattern printing. The distribution of

cell numbers is found to have a binomial form, which consistent with a

uniform distribution of cells inside the medium in the reservoir.

For pattern printing, L929 fibroblast cells were delivered by using a 60 µm

nozzle. Printed cells successfully kept their patterns in the crosslinked gel

made from 1.0% (w/v) alginate and 0.5% (w/v) calcium chloride. However, it

was found that the cells failed to adhere to alginate. On the other hand, cells

dispensed onto collagen gel were found to successfully maintain their viability,

adhere to the gel, spread and proliferate, forming a denser pattern. However,

unlike the crosslinked calcium-alginate which can immobilize cells quite

rapidly, cell adhesion to collagen needs a relatively long time to get

established. Therefore, some of the printed cells were slightly moved from

their initial position when the sample was disturbed, by the addition of fresh

medium or unintended shaking of the sample, which will reduce the resolution

of the printing. The smallest nozzle, with orifice diameter of 36 µm, was not

used for pattern printing, due to issues concerning the reliability of the printing

process, as it can easily get clogged.

List of Tables

x

List of Tables

Table 2-1: The minimum actuation pressure for droplet generation in DOD

inkjet devices [58]. ............................................................................................ 37

List of Figures

xi

List of Figures

Fig. 1.1: A typical flow diagram of photolithograph-based and inkjet printing

based process. ..................................................................................................... 3

Fig. 2.1: Layout of the different inkjet printing technologies. .......................... 11

Fig. 2.2: A Binary-Deflection continuous inkjet system. ................................. 13

Fig. 2.3: A Multilevel-Deflection continuous inkjet system............................. 13

Fig. 2.4: Droplets generated from a continuous inkjet system with multi-

nozzles............................................................................................................... 14

Fig. 2.5: Schematic of the DOD inkjet printing process. .................................. 15

Fig. 2.6: Droplet formation process within the ink chamber of a thermal inkjet

device. ............................................................................................................... 16

Fig. 2.7: Roof-shooter Thermal inkjet. ............................................................. 16

Fig. 2.8: Side-shooter Thermal inkjet. .............................................................. 17

Fig. 2.9: Schematic of the squeeze-mode inkjet. .............................................. 17

Fig. 2.10: Schematic of the bend-mode inkjet. ................................................. 18

Fig. 2.11: Schematic of the push-mode inkjet. ................................................. 19

Fig. 2.12: Schematic of the shear-mode inkjet. ................................................ 19

Fig. 2.13: Jet straightness error in both X and Y directions for Spectra SX-128

printhead [42]. ................................................................................................... 26

Fig. 2.14: Two nozzles to show the effects of hydrophobic treatment. (a).

Nozzle without hydrophobic treatment. (b). Nozzle with hydrophobic

treatment. .......................................................................................................... 27

Fig. 2.15: Image showing profiles of dried droplets printed on hydrophobic

and hydrophilic surfaces [44]............................................................................ 28

Fig. 2.16: Distinct dried droplet patterns under different temperature [45]. ..... 29

Fig. 2.17: Examples of five typical inkjet-printed line morphologies. (a).

Individual droplets. (b). Scalloped line. (c). Uniform line. (d). Bulging line. (e).

Stacked coins. Droplet spacing decreases from left to right [46]. .................... 31

List of Figures

xii

Fig. 2.18: Schematic representation of wave propagation and reflection in a

squeeze-mode piezoelectric inkjet printhead. ................................................... 32

Fig. 2.19: Schematic representation the basic energy requirement for ejecting a

droplet. .............................................................................................................. 37

Fig. 2.20: Effects of pulse amplitude on droplet velocity and droplet volume

[60]. ................................................................................................................... 39

Fig. 2.21: Effects of pulse width on droplet velocity and droplet volume [60].40

Fig. 2.22: Effects of jetting frequency on droplet velocity and droplet volume

[62]. ................................................................................................................... 41

Fig. 2.23: Sequence of images of DOD droplet formation for water [63]. ....... 43

Fig. 2.24: Different kinds of commercial printheads. ....................................... 45

Fig. 2.25: Schematic of the construction of a piezoelectric squeeze mode DOD

printhead. .......................................................................................................... 46

Fig. 2.26: Ejection nozzle orifice cross section requirements. ......................... 47

Fig. 2.27: KOH etching for a (100) silicon wafer. (a). Slice orientations for

silicon material. (b). Slice orientations shown in a plan view of a (100) silicon

wafer. Etching process proceeds downward until (111) planes are reached. (c).

“A-A” cross-section view. ................................................................................ 49

Fig. 2.28: Nozzle fabricated by silicon micromachining method comprising

KOH etching and Deep Reactive Ion Etching. (a). Plan view of the etched

wafer. (b). “A-A” cross-section view of the etched wafer. ............................... 50

Fig. 2.29: Schematic of photolithographically predefined inkjet printing. (a).

Schematic diagram of high-resolution inkjet printing onto a prepatterned

substrate. (b). AFM showing accurate alignment of inkjet-printed PEDOT/PSS

source and drain electrodes separated by a repelling polyimide (PI) line with L

= 5 µm. [20] ...................................................................................................... 53

Fig. 2.30: Schematic of pulse waveforms used for driving the inkjet printhead.

(a). A uni-polar waveform. (b). A bi-polar waveform. (c). The new waveform

for small droplet generation. [21] ..................................................................... 56

Fig. 2.31: (a) – (c) Images showing appearance and disappearance of a tongue

and formation of droplet with a diameter similar to that of the nozzle. (d) – (f)

Images showing formation of a droplet with a diameter much small than that

of the nozzle orifice. [21] .................................................................................. 56

Fig. 2.32: Schematic of an electrohydrodynamic jet system. [86] .................... 58

List of Figures

xiii

Fig. 2.33: Time-lapse images of the pulsating Taylor cone with the four stages

of the complete jetting cycle. Each frame is an average of 100 exposures with

the same delay. [89] .......................................................................................... 60

Fig. 2.34: High-resolution e-jet printing with printed feature size smaller than

1 µm. [86] ......................................................................................................... 61

Fig. 2.35: Printed cells. (a). 3-D tube structure made from printed cells. The

image shows an inner layer of human umbilical endothelial cells (green) and

an outer layer of human aortic smooth muscle cells (red). (b). Printed yeast

patterns after 3 days of culture. [2] ................................................................... 63

Fig. 2.36: 3D scaffold and the cells seeded into it. (a). A 3D scaffold

fabricated by rapid prototyping method. (b). Big view of the scaffold shown in

(a). (c). Human fibroblast cells seeded into a 3D scaffold, after 18 days of

culture. [121] ..................................................................................................... 66

Fig. 2.37: Fabrication of a scaffold by 3D plotting. (a). One layer. (b). Two

layers. [122] ...................................................................................................... 66

Fig. 2.38: Schematic diagram of organ printing. [138] ..................................... 68

Fig. 3.1: The novel printhead. (a) Schematic showing of the design (out of

proportion). (b) A self-fabricated printhead following the novel design. ........ 76

Fig. 3.2: Schematic showing the fabrication of the printhead chamber: (a) PET

tube before shrink. (b) Teflon tube before etching. (c) The steel tube used as a

mould during heating of PET. (d) PET tube after shrink. (e) Teflon tube after

etching. (f) Piezoelectric tube. (g) Shrunken PET tube bonded to the

piezoelectric tube. ............................................................................................. 77

Fig. 3.3: Schematic showing the design of the printhead housing and the

nozzle adaptor. .................................................................................................. 78

Fig. 3.4: Fabrication of a glass nozzle by heating and pulling glass tubing. (a)

Drawing of the glass tubing heating system (out of proportion). (b) Glass

tubing containing a hollow cone with a closed end. (c) A 50 µm orifice

fabricated by polishing the end of the tubing showing in (b). .......................... 79

Fig. 3.5: Fabricating glass nozzle by heating and pulling 1.0 mm glass

capillary with a micropipette puller. (a). The P-97 Flaming/Brown type

micropipette puller. (b). Heating the capillary. (c). Hit the sharp tip to from an

orifice. ............................................................................................................... 80

Fig. 3.6: Different shapes of tips fabricated by the micropipette puller. (a). A

too “sharp” tip. (b). A tip with a moderate converging shape. ......................... 81

Fig. 3.7: A 13-micron-tip fabricated by the micropipette puller. ...................... 82

Fig. 3.8: Inkjet printhead nozzles fabricated from glass tube. .......................... 82

List of Figures

xiv

Fig. 3.9: Schematic showing of the drop-on-demand inkjet printing system

used in the experiment. ..................................................................................... 85

Fig. 3.10: Image sequences showing the formation of a 50 µm droplet from a

36 µm inkjet nozzle. The times shown are 0, 144, 322, 367, 389, 400, 522 and

1122 µs relative to the first frame. The droplet velocity is here determined to

be 0.69 m/s. ....................................................................................................... 85

Fig. 3.11: Schematic showing of the uni-polar pulse waveform. ..................... 86

Fig. 3.12: Measured viscosities for different concentrations of sodium alginate

solutions. Measurement at 20 ˚C. ..................................................................... 87

Fig. 3.13: Threshold voltages for PET-based printhead (–○–), PTFE-based

printhead (–*–) and glass-based printhead (–■–). Nozzle diameter is 119 µm.89

Fig. 3.14: Effects of pulse width on droplet velocity and droplet size. The

pulse amplitude is 50 V. Nozzle diameter is 119 µm. ...................................... 91

Fig. 3.15: Effects of pulse amplitude on droplet velocity and droplet size. The

pulse width is 100 µs. Nozzle diameter is 119 µm. ......................................... 94

Fig. 3.16: Effects of nozzle size on droplet diameter. (–*–) denotes the

diameters of the smallest single droplets can be generated; (–■–) denotes the

diameters of the biggest single droplets can be generated; (–▲–) denotes the

diameters of the biggest droplets which can be generated using the maximum

voltage. .............................................................................................................. 97

Fig. 3.17: Repeatability test of the PET-based printhead. Nozzle diameter is

119 µm. ............................................................................................................. 98

Fig. 3.18: Effects of jetting frequency on droplet velocity and droplet size. The

pulse width is 100 µs. The pulse amplitude is 30 V. Nozzle diameter is 119

µm. .................................................................................................................... 99

Fig. 3.19: Threshold voltages for sodium alginate solutions of concentrations

from 0.2% to 2.8% (w/v). ............................................................................... 102

Fig. 3.20: Schematic showing of drop formation for 2.2% SA solutions. ...... 103

Fig. 4.1: Jet formation observed just after impact of the tube with a solid wall

when the free surface is initially deformed with a meniscus [168]. ............... 107

Fig. 4.2: A 93 µm jet with a velocity of 7 m/s. The diameter of the orifice

is 150 µm. Liquid used is 70% aqueous glycerin (w/w) solution.

Printing parameters: bi-polar piezo-driving signal with tdwell and techo equal to

700 µs; driving pulse amplitude equals to 140 V. Negative pressure inside the

reservoir is -2.2 kPa relative to the atmospheric pressure. Images were taken at

a frame rate of 8 kfps. Ambient temperature is 25 ˚C. ................................... 109

List of Figures

xv

Fig. 4.3: The 150 µm nozzle used for fine jetting experiments. The scale bar is

2 mm. This image was taken when the nozzle was placed inside a 60%

aqueous glycerin (w/w) solution, which had an index of refraction similar to

that of the glass. .............................................................................................. 110

Fig. 4.4: An 8 µm jet with a velocity of 29 m/s. is 150 µm. The liquid

used is 70% aqueous glycerin (w/w) solution. Printing parameters: bi-polar

piezo-driving signal with tdwell and techo equal to 700 µs; driving pulse

amplitude equals to 140 V. The negative pressure inside the reservoir is -2.3

kPa relative to the atmospheric pressure. Images were taken at a frame rate of

165 kfps. Ambient temperature is 25 ˚C. The scale bar is 500 µm. ............... 110

Fig. 4.5: A 16 µm jet with a velocity of 35 m/s. is 150 µm. The liquid


piezo-driving signal with tdwell and techo equal to 700 µs; driving pulse

amplitude equals to 140 V. Negative pressure inside the reservoir is -2.3 kPa

relative to the atmospheric pressure. Images were taken at a frame rate of 16

kfps. Ambient temperature is 25 ˚C. ............................................................... 111



piezo-driving signal with 450 µs tdwell and 70 µs techo; driving pulse amplitude

equals to 140 V. The negative pressure inside the reservoir is -2.3 kPa relative

to the atmospheric pressure. Images were taken at a frame rate of 27 kfps.

Ambient temperature is 25 ˚C. The scale bar is 500 µm. ............................... 112


used is water. Printing parameters: bi-polar piezo-driving signal with 700 µs

tdwell and 700 µs techo; driving pulse amplitude equals to 140 V. The negative

pressure inside the reservoir is -2.3 kPa relative to the atmospheric pressure.

Images were taken at a frame rate of 330 kfps. Ambient temperature is 25 ˚C.

The scale bar is 500 µm. (a). The time interval between successive frames, dt,

equals to 9.09 µs. (b). dt equals to 3.03 µs. (c). dt equals to 9.09 µs. ............. 114

Fig. 4.8: Schematic showing the free surface shapes...................................... 116


used is water. Printing parameters: bi-polar piezo-driving signal with 500 µs

tdwell and 500 µs techo; driving pulse amplitude equals to 140 V. Negative


Images were taken at a frame rate of 330 kfps. The numbers of the frames

shown in the figure are n = 1, 4, 7 …… 52. Ambient temperature is 25 ˚C. The

scale bar is 500 µm. ........................................................................................ 117

Fig. 4.10: Images showing jetting produced when no coalescence happens

between the two cavities. is 150 µm. The liquid used is 70% aqueous

glycerin (w/w) solution. Printing parameters: bi-polar piezo-driving signal

with 550 µs tdwell and 550 µs techo; driving pulse amplitude equals to 140 V.

The negative pressure inside the reservoir is -2.3 kPa relative to the

atmospheric pressure. Images were taken at a frame rate of 330 kfps. Ambient

List of Figures

xvi

temperature is 25 ˚C. The scale bar is 500 µm. (a). dt equals to 6.06 µs. (b). dt

equals to 3.03 µs. (c). dt equals to 9.09 µs. (c). dt equals to 6.06 µs. ............. 118

Fig. 4.11: Images showing the cavity jet pierces the thin liquid film. is

150 µm. The liquid used is 30% aqueous glycerin (w/w) solution. Printing

parameters: bi-polar piezo-driving signal with 750 µs tdwell and 750 µs techo;

driving pulse amplitude equals to 140 V. The negative pressure inside the


a frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 200

µm. (a). dt equals to 6.06 µs. (b). dt equals to 3.03 µs. .................................. 120

Fig. 4.12: Images showing the cavity jet fails to pierces the cavity. is

150 µm. The liquid used is 85% aqueous glycerin (w/w) solution. Printing

parameters: bi-polar piezo-driving signal with 650 µs tdwell and 650 µs techo;

driving pulse amplitude equals to 140 V. The negative pressure inside the


a frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 200

µm. (a). dt equals to 18.18 µs. (b). dt equals to 3.03 µs. (c). dt equals to 15.15

µs. .................................................................................................................... 121

Fig. 4.13: A thin liquid thread generated during the jetting. is 150 µm.

The liquid used is 70% aqueous glycerin (w/w) solution. Printing parameters:

bi-polar piezo-driving signal with 550 µs tdwell and 550 µs techo; driving pulse



330 kfps. Ambient temperature is 25 ˚C. The scale bar is 200 µm. (a). dt

equals to 3.03 µs. (b). dt equals to 12.12 µs. (c). dt equals to 6.06 µs. ........... 122

Fig. 4.14: Images showing the interaction between the piezo-generated cavity

and the preexisting bubble inside the nozzle. is 150 µm. The liquid used

is 75% aqueous glycerin (w/w) solution. Printing parameters: bi-polar piezo-

driving signal with 550 µs tdwell and 550 µs techo; driving pulse amplitude

equals to 140 V. The negative pressure inside the reservoir is -2.3 kPa relative

to the atmospheric pressure. Images were taken at a frame rate of 330 kfps.

Ambient temperature is 25 ˚C. The scale bar is 500 µm. dt equals to 6.06 µs.123

Fig. 4.15: Surfaces collapse jet upward into the nozzle. is 150 µm. The

liquid used is 85% aqueous glycerin (w/w) solution. Printing parameters: bi-

polar piezo-driving signal with 650 µs tdwell and 650 µs techo; driving pulse



165 kfps. Ambient temperature is 25 ˚C. (a). The scale bar is 1 mm. dt equals

to 24.24 µs. (b). The scale bar is 1 mm. dt equals to 18.18 µs. (c). The scale

bar is 500 µm. dt equals to 12.12 µs. .............................................................. 125

Fig. 4.16: Surfaces collapse jets. is 150 µm. The liquid used is 50%

aqueous glycerin (w/w) solution. Printing parameters: bi-polar piezo-driving

signal with 550 µs tdwell and 550 µs techo; driving pulse amplitude equals to 140

V. The negative pressure inside the reservoir is -2.3 kPa relative to the


List of Figures

xvii

temperature is 25 ˚C. The scale bar is 200 µm. Image for frame number n = 1,

3, 5 …… 13, 15............................................................................................... 126

Fig. 4.17: Jetting velocities obtained for different concentration of aqueous

glycerin solutions (w/w): 0%, 10%, 30%, 50%, 70%, 75%, 80%, and 85%. . 126

Fig. 4.18: The fastest jet observed in the experiment: a 9 µm jet with a velocity

of about 100 m/s. is 150 µm. The liquid used is 50% aqueous glycerin

(w/w) solution. Printing parameters: bi-polar piezo-driving signal with 550 µs

tdwell and 550 µs techo; driving pulse amplitude equals to 140 V. The negative



The scale bar is 200 µm. Time interval between frames is dt = 3.03 µs. ....... 128

Fig. 4.19: Images showing the relationship between jet velocity and jet

diameter. Jets belong to type II. is 150 µm. The liquid used is 70%

aqueous glycerin (w/w) solution. Printing parameters: bi-polar piezo-driving

signal with 550 µs tdwell and 550 µs techo; driving pulse amplitude equals to 140

V. The negative pressure inside the reservoir is -2.3 kPa relative to the


temperature is 25 ˚C. The scale bar is 200 µm. (a). A 1 µm jet with a velocity

of 66 m/s. (b). A 3 µm jet with a velocity of 51 m/s. (c). A 10 µm jet with a

velocity of 15 m/s............................................................................................ 129

Fig. 4.20: Images showing the relationship between jet velocity and jet

diameter. Data collected for both Jet I and Jet II. is 150 µm. Liquid used

is 0%, 10%, 30%, 50%, 70%, 75%, 80% and 85% aqueous glycerin (w/w)

solutions. Printing parameters: bi-polar piezo-driving signal; driving pulse

amplitude equals to 140 V. Ambient temperature is 25 ˚C............................. 130

Fig. 5.1: Schematic showing the DOD setup for cell printing experiment. .... 137

Fig. 5.2: Images taken by using the high-speed-video camera. (a). Image

sequence showing the formation of a 160 µm droplet from a 119 µm nozzle,

taken at a frame rate of 8,000 fps, giving time between frames of 125 µs.

Liquid used was 1.0% (w/v) aqueous solution of sodium alginate. Drop

velocity is 0.74 m/s. (b). Images showing cell motion inside the nozzle. Nozzle

opening diameter is 119 µm. ........................................................................... 137

Fig. 5.3: Graph showing influence of excitation pulse on droplet velocity. The

orifice diameters of the nozzles used were 36, 81 and 119 µm. ..................... 140

Fig. 5.4: Graph showing influence of excitation pulse voltage on droplet

diameter. The orifice diameters of the nozzles used were 36, 81 and 119 µm.141

Fig. 5.5: Graph showing a 95% survival rate of L929 rat fibroblast cells

stained with Calcein AM and Ethidium homodimer-1. Printed with an

excitation pulse amplitude of 116 V, at a frequency of 1.5 kHz, with rising and

falling times of 3 µs. The orifice used was 119 µm. ....................................... 142

List of Figures

xviii

Fig. 5.6: Mean cell survival rate with respect to excitation pulse amplitude for

the samples printed through the 36 µm orifice, with excitation pulse amplitude

from 60 V to 130 V, at a frequency of 1.5 kHz, with rising and falling times of

3 µs. Error bars show the standard error from 5 replicates. ............................ 143

Fig. 5.7: Graph showing the mean cell survival rate against excitation pulse

amplitude. Samples printed through orifices with the diameter of 36, 81 and

119 µm, with excitation pulse amplitude from 52 to 140 V, at frequency of 1.5

kHz, with rising and falling times of 3 µs. Each cell survival rate data was the

average value from 5 replicates. ..................................................................... 144

Fig. 5.8: Graph showing percentage of cell death against the mean shear rate.

Samples printed through orifices with the diameter of 36 µm, 81 µm and 119

µm. Each cell death rate data was the average value from 5 replicates. ......... 145

Fig. 5.9: Droplets printed onto a dry substrate from a suspension with a

concentration of 2×106 cells per ml. Each droplet contains 1 to 5 cells. The

orifice diameter of the nozzle used was 60 µm. ............................................. 147

Fig. 5.10: Graph showing the probability density distribution of the number of

cells in each droplet. For a range of different average cell concentration in the

cell medium, from dN = 0.5, 1.0, 1.5 … 3.0 cells per droplet. ...................... 149

Fig. 5.11: Optical micrographs of L929 rat fibroblast cells after 5 days in

culture following printing. Cell division can be observed (indicated by green

circle) apparently. ........................................................................................... 150

Fig. 5.12: Images of printed cells. (a). Cells inside dried droplet residues. The

scale bar is 50 µm. (b). Schematic showing the measurement of the radial

location of each cell, away from the center of the dried droplet residue. ....... 150

Fig. 5.13: Graph showing the probability density distribution of the number of

cells in each droplet. The (□) stands for the experimental results and (--+--)

stands for the values calculated from eq. 5.3. Determined from microscope

counting of cells in 800 droplets, which were dispensed within the first 4

minutes. ........................................................................................................... 151

Fig. 5.14: Graph showing the probability of cell location within the dried

droplet splatter. The “radius” is the distance from the center of the cell to the

center of the dried droplet. The “Radius” is the radius of the dried droplet.

“Rcell” is the radius of the round-shaped L929 rat fibroblast cells, which has a

value of approximately 10 µm. ....................................................................... 152

Fig. 5.15: Graph showing the average number of cells per droplet vs. time

from start of printing. Printing was carried out continuously over a period of

2.5 hours, at 120 Hz driving frequency........................................................... 153

Fig. 5.16: Image showing cells printed onto a dry Petri-dish, forming an

“NUS” pattern. Each droplet contains 2 to 6 cells. The orifice diameter of the

nozzle used was 60 µm. .................................................................................. 154

List of Figures

xix

Fig. 5.17: Image showing a continuous line of overlapping droplets with

around 6 to 8 cells per droplet in the crosslinked gel. The orifice diameter of

the nozzle used was 60 µm. ............................................................................ 154

Fig. 5.18: Image showing live cells printed onto a collagen gel, forming an

“NUS” pattern. The orifice diameter of the nozzle used was 60 µm. Picture

taken 5 day after printing. ............................................................................... 155

List of Symbols

xx

List of Symbols

f Jetting frequency

c Sound speed

v Droplet velocity

d Droplet diameter

Dynamic viscosity of liquid

Surface tension of liquid

Density of fluid

Piezoelectric strain constant

Kinematic viscosity of liquid

N Average cell concentration per unit volume

Droplet volume

Average number of cells per droplet volume

Droplet impact velocity

Contact angle

Surface tension force at the liquid-gas interface

Chapter 1: Introduction

1 of 176

1. INTRODUCTION

1.1 Background

For environmental conservation and the realization of a sustainable society, it

is necessary that industrial manufacturing processes undergo a transformation

with reduction of environmental impact. From this viewpoint, additive

manufacturing technologies have attracted considerable attention because they

have the potential to greatly reduce ecological footprints as well as the energy

consumed in manufacturing. An additive manufacturing process is one

whereby a product is made by adding successive layers of material onto a

substrate. Examples are electron beam melting, laser sintering, aerosol jet

printing and inkjet printing. Rapid Prototyping (RP) is the name generally

given to the various additive processes. Besides the above mentioned

advantage of environmental benignity, additive manufacturing process is also

a low cost production method for reducing the material wastage, especially for

the specialty polymers and precious metals.

Drop-on-Demand Inkjet Printing (DOD IJP) is an additive manufacturing

process, a data-driven process that patterns directly onto the substrate with

ejected droplets. It is capable of precise deposition of picoliter volumes (down

to 2 pL, 15 m in diameter) of liquids at high speed (up to 60 kHz [1]) and

accuracy (< 5 m) on a target surface, even onto non-planar surface. Due to its

advantages in high resolution, automation, low cost, non-contact, flexible,

environmental benignity and ease of material handling, the application of


2 of 176

DOD inkjet printing technology has been expanded from conventional graphic

printing to new areas, such as organ printing, displays, integrated circuits (ICs),

solar cells, memory devices, optical devices, MEMS and drug delivery.

One of the aims of tissue engineering is to position cells into 3-D structures

and arrange them in a specific pattern. The generation of such structures

forms the basis of tissue regeneration and possibly, organ building [2]. Inkjet

printing is a suitable candidate for this purpose. It has been used successfully

in a similar manner for automated rapid prototyping technology which

precisely positions droplets onto a substrate. To date, many different cell types

have been printed successfully by different printing methods and their viability

has been verified [3-10]. The power of inkjet printing lies in its ability to

deliver picoliter volumes of materials at high speed and accuracy on a target

interface (probably non-planar surface), and to deliver active substances to a

developing structure in timing sequence. By using different cell types as

different bio-inks, and delivering them to exact positions to mimic tissue

structures of the original tissue, inkjet printing offers a possible solution for

building whole structures such as bone, cornea, ligament, cartilage etc, to

solve the organ transplantation crisis.

With improving living standards, requirements for low-power, fast response

time, lightweight, wide viewing angle and portable communication devices are

rising and galvanizing the display industry to loot at a new technology known

as polymer light-emitting-diode (PLED) display. Monochromatic displays can

be prepared by spin-coating; however, to fabricate a full-color PLED flat-


3 of 176

panel display (FPD), a micro-patterned array of red, green, and blue PLED

subpixels must be fabricated on the display backplane. This requires the three

differently colored electroluminescent polymers to be deposited onto the exact

position of the substrate [11, 12]. The spin-coating technique is clearly not

suitable for such displays. Subtractive patterning, such as the

photolithographic technique, is also not appropriate for such task due to its

high cost, and complicated process as a multi-stage approach. Among all the

manufacturing processes, inkjet printing has been proved to be the most

promise technique for full-color PLED displays fabrication, and PLED devices

have been demonstrated by plenty of companies such as Seiko-Epson, Philips,

CDT, DuPont, Samsung SDI, TM (Toshiba-Matsushita) Display and Delta.

Fig. 1.1: A typical flow diagram of photolithograph-based and inkjet printing based

process.

Applying inkjet printing technology to electronics patterning is quite

straightforward, as material can be deposited on-demand, which reduces

material wastage. It is also well-known that a conventional silicon patterning


4 of 176

process usually involves photolithography and etching processes (either

reactive ion etching or anisotropic wet chemical etching), which consist of

many sub-processes and lead to long processing time and high cost, as shown

in Fig. 1.1; all these complicated processes are avoided in inkjet printing as it

is a non-lithographic patterning method. Besides saving the cost of lithography

masks and materials, DOD inkjet printing also has many other advantages.

Firstly, as a low temperature process, micro-patterning process can be even

performed on paper or plastics, which makes it well suited to roll-to-roll

fabrication and makes it especially attractive for fabricating large-area, ultra-

low cost electronic circuits on flexible substrates [13, 14, 15]. Secondly,

applications of the above photolithographic patterning and etching processes

to polymer multilayer structures is difficult because of the plasma-induced

degradation of electroactive polymers and the lack of suitable anisotropic

etching techniques for polymers [16]. However, inkjet printing can handle a

wide range of materials including solution-based materials, suspended nano-

particles and polymers; it also allows the use of inviscid ink without added

binders [17]. This feature makes it a possible technique for low-cost

fabrication of solution-processible polymer field-effect electronics devices

[18]. Thirdly, inkjet printing is a data-driven process that can directly transfer

computer-aided designs into device patterns, which can greatly save the

process time and accommodate customization. To conclude, electronics

fabricated by direct inkjet printing of functional electronic materials has

gained significant interest as an alternative to conventional silicon integrated

circuit (IC) process.


5 of 176

1.2 Challenges

As discussed in above section, because of its unique advantages, DOD inkjet

printing has emerged as an attractive patterning technique for a variety of new

areas in the last two decades. Accordingly, the dispensed liquids have been

expanded from the conventional pigmented ink (or standard dye-based ink) to

polymers, gels, cell ink or other materials which often have higher viscosities

or even contain large particles or cells. For simplicity, the word “ink” is still

used to represent the liquid to be dispensed. Ink viscosity is the most crucial

parameter which will affect printing. When the actuator is activated, energy

goes into kinetic energy, viscous flow and surface tension of the free-surface

flow. Viscous dissipation causes partial energy loss in the printhead. As a

result, ink viscosity must be low enough to ensure the success of droplet

dispensing. For most of the commercial inkjet printheads supplied by

companies like Microdrop, Microfab, Dimatix and XAAR, only liquids with

viscosities lower than 20 cps [12] can be consistently dispensed. Fluids with

even higher viscosities have to be diluted before printing or warmed up during

the printing, which will either adversely affect the properties of the liquids or

lead to long processing times in printing.

Another major concern in inkjet printhead design is the “first drop problem”,

which is the clogging of nozzles by dried ink at the nozzle tip. Especially,

when inkjet printing is applied to the above new areas, inks containing

particles, or even cells, can easily block the nozzle orifice, resulting in time-

consuming nozzle cleaning or even damage of the entire printhead. To solve


6 of 176

the problem, the easiest way is to use a nozzle with a bigger orifice, as bigger

orifices are less likely to clog. However, this is often not desirable in inkjet

printing as bigger nozzles result in bigger droplets and lower printing

resolution. Especially, for applications such as fabricating organic transistor

circuits or MEMS devices, the resolution of current inkjet printing is still too

low (normally limited to 20 µm by droplet size and the spreading of the

droplet on the substrate [19, 20]) and droplet size needs to be further reduced.

Besides reducing the nozzle size, when using piezoelectric-based DOD inkjet

printhead, it has been proved that smaller droplets could be produced by

judiciously controlling the piezoelectric parameters [21, 24, 25]. These studies

reveal a possible way to alleviate the nozzle clogging problem without

sacrificing printing resolution. However, these methods only work over a

limited range of Ohnesorge numbers and their effects are also limited: the

diameter of the dispensed droplets can be only reduced to a maximum of 60 %

of the orifice diameter. Consequently, reducing nozzle size seems the only

efficient way to reduce droplet size, to fulfill the resolution requirement by the

new applications of inkjet printing, such as fabrication of organic transistor

circuits or MEMS devices. As can be foreseen, the clogging problem would

become even worse during printing.

The poor printability and nozzle clogging may result in unreliable or even

failed dispensing and thus impose tremendous challenges on the printhead

design and printing process.


7 of 176

1.3 Objectives

Main objectives of this research are:

To study drop generation conditions and ejection nozzle requirements

in DOD inkjet printing. Two methods for fabricating microscale inkjet

nozzles, based on micro-pipette fabrication technology and silicon

micro-machined technology, will be proposed and tested.

To design and fabricate a new type of PET/PTFE-based piezoelectric

squeeze mode inkjet printhead. The new printhead should have the

ability to dispense liquids with much higher viscosities (> 100 cps).

The new printhead should also have interchangeable nozzle design, so

the clogged nozzle can be easily removed and cleaned. Especially, the

damaged nozzle can be easily changed, avoiding the destruction of the

whole printhead assembly.

To characterize the in-house-developed printhead: investigating the

printability and the printing repeatability of the new printhead by

comparing it with the conventional glass-based printheads;

investigating the effects of printing parameters (pulse amplitude, pulse

width, nozzle size, jetting frequency etc.) on droplet velocity and

droplet diameter; optimizing the printhead design to improve the

maximum jetting frequency.


8 of 176

To investigate the efficiency of three different methodologies on

generation of microscale droplets: reducing the droplet size by directly

reducing the nozzle size down to 1 to 2 microns; carefully controlling

the piezoelectric waveforms to generate droplets smaller than the

nozzle size; generating much smaller droplets or fine jet by combining

DOD inkjet printing with the conventional electrospinning technique.

To carry out the cell printing experiments. Investigate the survivability

of the cells subjected to the large stresses during the printing process.

Fibroblast cells will be printed onto different substrates (alginate,

collagen etc) and cultured over a period of days to verify their long-

term viability. Pattern printing, cell agglomeration in the cell ink, cell

number in each printed droplet and cell location inside the dried-

droplet will also be studied.

1.4 Organization

The layout of this thesis is organized as follows:

Chapter 2 presents an essential introductory knowledge on the inkjet

printing technology, which includes the classification of different DOD

inkjet printing methods and their work principles, conditions for

dispensing a droplet from an inkjet nozzle, different ejection nozzle

fabrication methods and different criteria for evaluating printing


9 of 176

quality. It also presents the state-of-the-art work in the areas of

generating ultra-small droplets and cell printing.

Based on our theoretical study, a PET/PTFE-based piezoelectric

squeeze-mode DOD inkjet printhead with interchangeable nozzles, has

been designed and fabricated, which will be discussed in Chapter 3.

The detailed printhead chamber design and the ejection nozzle

fabrication process will be given. The advantages of the in-house-

developed printhead, as well as its characterizations will also be

discussed in detailed.

In Chapter 4, the fine jet generated in DOD inkjet printing will also be

systematically studied, with the help of an ultra-high-speed, high

space-resolution video camera system.

Chapter 5 presents the results of the cell printing experiments. The

effects of shear stresses on cell survivability, the long-term viability of

the cells printed onto different substrates (coated by alginate or

collagen), and the results of pattern printing will all be discussed in

detailed.

Chapter 7 outlines future working directions that could further improve

the printhead resolution and the maximum jetting frequency, based on

the theoretical and experimental work presented in this dissertation.

Chapter 2: Literature Review

10 of 176

2. LITERATURE REVIEW

2.1 Introduction to Inkjet Printing

Inkjet printing is a contact-free dot-matrix printing technique in which an

image is created by directly jetting ink droplets onto specific locations on a

substrate [26]. The concept of inkjet printing can trace its history to the 19th

century and the inkjet printing technology was first developed in the early

1950s. Inkjet printers that capable of reproducing digital images generated by

computers were developed in the late 1970s, mainly by Hewlett-Packard,

Epson and Canon. The booming of the personal computer industry in 1980s

has led to a substantial growth of the printer market and nowadays personal

printer is present in almost every office and home. Inkjet printing technology

is developing at a rapid pace. It has been expanded from conventional graphic

printing to various applications, such as organ printing, displays, integrated

circuits (ICs), optical devices, MEMS and drug delivery.

2.1.1 Classification of Inkjet Printing Techniques

Inkjet printing technology has been developed in a wide variety of ways. In

Fig. 2.1, the inkjet tree structure shows a layout for most of the better known

inkjet printing techniques and some of the corresponding adopters. As can be

seen, there are two categories of inkjet printing technology: Continuous inkjet

printing and Drop-on-Demand inkjet printing.


11 of 176

Fig. 2.1: Layout of the different inkjet printing technologies.

2.1.1.1 Continuous Inkjet Printing

The earliest inkjet devices operated in a continuous mode. The idea was first

patented by Lord Kelvin in 1867 and the first commercial model was

introduced by Siemens in 1951. In this technique, a continuous jet of the liquid

ink is formed by applying pressure to the ink chamber with a small orifice at

one end. A fluid jet is inherently unstable and will break up into droplets,

which is entirely a consequence of the surface tension effects. This

phenomenon was firstly noted by Savart in 1833 and described mathematically

by Lord Rayleigh [27]. If surface tension force is the only force acting on the


12 of 176

free surface of the jet, it will break up into droplets of varying size and

velocity; when a periodic perturbation of an appropriate frequency is applied

to the liquid, typically using a piezoelectric transducer, the jet will break up

into droplets of uniform size and velocity. The droplets separate from the jet in

the presence of a properly-controlled electrostatic field which generated by an

electrode that surrounds the region where break-off occurs. As a result, an

electric charge can be induced on the drops selectively. Subsequently, when

the droplets pass through another electric filed, the charged droplets are

directed to their desired location on the substrate to form an image; those

uncharged droplets will drift into a catcher for recirculation. Continuous inkjet

can be classified into binary deflection or multilevel deflection according to

the drop deflection methodology, as can be seen in Fig. 2.1.

Fig. 2.2 and Fig. 2.3 schematically show streams of droplets generated from

binary deflection and multilevel deflection mode continuous inkjet,

respectively. A piezoelectric transducer is used to generate a periodic

perturbation onto the jet. In the both modes, the charged droplets are directed

to deposit onto the substrate, while in the multilevel, the charged droplets are

allowed to deposit onto the substrate at different levels. By using the

multilevel deflection system, a small image swath can be created by a single

nozzle. Fig. 2.4 shows droplets generated by a continuous inkjet system with

multi-nozzles. Continuous inkjet is widely used in the industrial coding,

marking, and labeling markets [26]. Extensive studies, both theoretical and

experimental, have been conducted to analysis different continuous inkjet

systems, especially the process of disturbance growth on the jet stream which


13 of 176

leads to droplet formation. Typically the droplets generated by a continuous

inkjet system have a diameter of approximately twice of the orifice diameter.

Droplets sizes range from 20 µm to 500 µm can be generated at rates of up to

1 MHz by continuous inkjet.

Fig. 2.2: A Binary-Deflection continuous inkjet system.

Fig. 2.3: A Multilevel-Deflection continuous inkjet system.


14 of 176

Fig. 2.4: Droplets generated from a continuous inkjet system with multi-nozzles.

The major advantage of continuous inkjet is that it can generate ink droplets

with very high velocity, which can reach to 50 m/s. This feature allows for the

usage of a relatively long distance between printhead and substrate. It also

allows for rapid droplet formation rate, also known as high speed printing.

Another advantage of continuous inkjet is no waste of ink, due to droplet

recycling. Furthermore, since the jet is always in use, nozzle clogging can be

avoided in continuous inkjet. Therefore volatile solvents such as alcohol and

ketone can be employed to promote drying of droplets onto the substrate.

The major disadvantage of continuous inkjet is that the ink to be used must be

electrically conducting, to ensure that ink droplets can be charged and directed

to the desired location. Furthermore, due to ink recycling process, ink can be

contaminated.

2.1.1.2 Drop-on-Demand Inkjet Printing

Drop-on-Demand inkjet systems were developed in the 1970s, when different

actuation principles were utilized [28]. In this technique, ink droplets are


15 of 176

produced only when they are required. According to the mechanism used

during the droplet formation process, DOD inkjet can be categorized into four

major types: thermal mode, piezoelectric mode, electrostatic mode, and

acoustic mode, as can be seen from Fig. 2.1. Most of the DOD systems in the

market are using the thermal or the piezoelectric modes. Nevertheless, no

matter which mode is used, the basic principles of all these different inkjet

methods are similar: a transducer, normally a piezoelectric element or a

thermal heater, generates a pressure pulse into the ink and forces a droplet out

of the orifice, as schematically shown in Fig. 2.5. The only difference lies in

that the way how this pressure pulse is generated.

Fig. 2.5: Schematic of the DOD inkjet printing process.

The first thermal inkjet device was designed in 1977 by Canon engineer Ichiro

Endo. In this technique, when a droplet is required, a current pulse of less than

a few microseconds is produced and passes through a heating element located

nearby the nozzle. Heat is transferred from the heater to the ink, causing a

rapid vaporization of the ink to form a vapor bubble inside the ink chamber.

As the ink chamber volume is fixed, this instantaneous expansion of bubble

will cause a large pressure increase inside the chamber, propelling the ink out


16 of 176

of the nozzle. Simultaneously with the later bubble collapse, the pushed-out

ink column will break off from the nozzle and form a droplet, flying to the

substrate. The duration of the air bubble formation and collapse is less than 10

µs. Fig. 2.6 schematically shows the droplet formation process in a thermal

inkjet chamber. As the bubble collapses, a vacuum is created. The ink then

flow back into the chamber and recover to its equilibrium state, waiting the

next round of jetting. According to its configuration, thermal inkjet device can

be classified into a roof shooter or a side-shooter type. The orifice is located

on top of the heating element in a roof-shooter thermal inkjet, while it is

located on a side nearby the heating element, as shown in Fig. 2.7 and Fig. 2.8.

Most of the consumer inkjet printers designed by companies such as Hewlett-

Packard and Canon are in thermal bubble type.

Fig. 2.6: Droplet formation process within the ink chamber of a thermal inkjet device.

Fig. 2.7: Roof-shooter Thermal inkjet.


17 of 176

Fig. 2.8: Side-shooter Thermal inkjet.

The earliest piezoelectric inkjet printhead was designed by Zoltan in 1972. In

this technique, when a droplet is required, an electric pulse will be applied to a

piezoelectric element located behind the nozzle. Then the piezoelectric

element changes its shape, causing a pressure pulse inside the ink that

propelling a droplet from the nozzle. Depending on the deformation method of

the piezoelectric element used in the device, piezoelectric inkjet printing can

be classified into four categories: squeeze mode, bend mode, push mode and

shear mode.

Fig. 2.9: Schematic of the squeeze-mode inkjet.

In squeeze mode piezoelectric inkjet, a thin piezoelectric tube is tightly

attached onto a glass tube which with an orifice at one end, as shown in Fig.

2.9. The piezoelectric tube is radially polarized and is with electrodes on its

outer and inner surfaces. When a droplet is desired, an electrical pulse will be


18 of 176

applied to the piezoelectric transducer, the polarity is selected to cause a

contraction of the transducer. As a result, the glass tube as well as the ink will

also be squeezed, and a droplet of ink will be ejected from the nozzle. Squeeze

mode piezoelectric inkjet is implemented by companies, such as Siemens,

Microdrop and MicroFab.

Fig. 2.10: Schematic of the bend-mode inkjet.

Fig. 2.10 schematically shows a piezoelectric actuator operating in bend mode.

The device consists of an ink chamber with one side of it formed by a

conductive diaphragm. A piezoelectric plate is tightly bonded to the

diaphragm. When an electric pulse is applied, the piezoelectric element will

contract in the radial direction, causing the diaphragm to flex inwardly into the

ink chamber. This instantaneous motion of diaphragm will cause a large

pressure increase inside the chamber and forces a droplet to be jetted from the

orifice. Successful implementation of the bend mode piezoelectric inkjet can

be found in printheads from companies, such as Epson and Sharp.

In a push mode piezoelectric design as shown in Fig. 2.11, when the

piezoelectric rod expands in the horizontal direction, it pushes against the ink

to eject a droplet from the orifice. Similar as in the bend mode, a thin


19 of 176

diaphragm is incorporated between the piezoelectric element and the ink, to

prevent undesirable interaction between ink and the piezoelectric materials

[29]. Push mode inkjet is implemented by companies, such as Epson and

Trident.

Fig. 2.11: Schematic of the push-mode inkjet.

Fig. 2.12: Schematic of the shear-mode inkjet.

In all above 3 types of inkjet devices, the electric field generated between

electrodes is parallel with the polarization of the piezoelectric plate. However,

in the shear mode piezoelectric inkjet device, the imposed electric fields are

orthogonal to the polarization direction of the piezoelectric element [30]. As

schematically shown in Fig. 2.12, P denotes the polarization directions; the

electrodes are mounted on the different locations of the piezoelectric plate.

Therefore, the resulting shear motion of the transducer decreases the volume


20 of 176

of the ink chamber and pushes the ink, ejecting a droplet from the orifice. In

shear mode piezoelectric inkjet, since the piezoelectric transducer directly

forms one wall of the ink chamber, thus interaction between ink and

piezoelectric materials is typically inevitable. It is also one of the key

parameters of shear mode inkjet [26]. Successful implementation of the shear

mode piezoelectric inkjet can be found in printheads from companies, such as

Spectra and Xaar.

Most of the industrial inkjet printers and some of the consumer printers (those

produced by Epson) are designed in piezoelectric type.

As can be seen from above discussion, the DOD actuation principle eliminates

the need for droplet charging system, droplet deflection system and ink

recycling system, thus the whole jetting device is more compact as compared

to continuous inkjet device. Furthermore, wider range of inks can be used in

DOD system as droplet charging is not required. Finally, since ink

recirculation is avoided, thus ink contamination can be eliminated. Currently

the majority of interest in inkjet printing is in the DOD methods.

The main disadvantage of DOD inkjet is the clogging of jet nozzles. Clogging

may result from particles inside the ink, especially when pigment-based ink is

used. Therefore, fine filters must be adopted upstream from the nozzle to

prevent relatively big particles from reaching the nozzle. Furthermore, during

the off working state, a solid deposit in the nozzle will form due to dry of ink,

which will also lead to nozzle clogging. This is well known as the “first drop


21 of 176

problem” [31]. Another disadvantage of DOD inkjet is the lower droplet

velocity. Typically DOD inkjet produce droplet with velocity lower than 10

m/s. Finally, DOD inkjet also has the advantage of low jetting speed. When a

droplet is produced, acceleration of a mass of ink is always required, and this

acceleration is created only by the actuating signal itself. Thus the droplet

generation rate, as well as the printing speed is limited for DOD devices [28].

2.1.2 Advantages and Disadvantages of Inkjet Printing

As a Rapid Prototyping technique, inkjet is an additive manufacturing process.

It ejects droplets only when required and hence reduces the material wastage.

This implies a lower cost for the applications that requires expensive materials,

which is sponsored for conservation and the realization of a sustainable

society.

As compared to the traditional photolithography-based patterning process,

which consists of many sub-processes and leads to long processing time and

high cost, inkjet printing is much more compact. It avoids all those

complicated sub-processes. Furthermore, it also saves the cost for lithography

masks as well as the huge work for storing the hundreds of masks.

Inkjet printing is a data-driven direct-write process that can directly transfer

Computer Aided Design (CAD) into device patterns, which can greatly save

the process time and accommodate customization.


22 of 176

Inkjet printing offers the advantage of non-contact between the nozzle and the

substrate, thus there is no mechanical wear on the printed sample. The

possibility of cross-contamination is also reduced to a minimum.

To conclude, inkjet printing offers advantages in low cost, compact,

automation, non-contact, environmental benignity and ease of material

handling. It is a highly flexible technology that is able to accurately deposit

small volumes of materials in almost any pattern.

There are two main disadvantages in inkjet printing. Firstly, the anisotropic

nature of the inkjet process, due to the intrinsic pinhole nature of the deposited

ink, results in the uneven surface roughness of the printed features [32].

Furthermore, this film non-uniformity can also be produced by the inevitable

“coffee stain” effect, which arises from interaction of multiple effects of the

solvent drying process [33, 34, 35]. This disadvantage in film uniformity does

not exist in planar processing currently used in industry. Secondly, the

resolution of inkjet method is limited by the dispensed droplets size. The

minimum diameter of an inkjet droplet for a state of the art inkjet printhead is

around 10 µm. The resulting printing resolution is enough for document

printing use, but for nanotechnology use and many other industrial

applications, such resolution is not good enough and a more precise inkjet

device needs to be developed. Furthermore, the attainable feature size of a

component fabricated by inkjet printing is also affected by how the droplet

interacts with the substrate. Thus the impact behaviour of the droplets onto a

substrate also needs to be carefully studied.


23 of 176

2.1.3 Printing System Evaluation

2.1.3.1 Print Resolution

Print resolution is normally measured by the number of individual dots that

can be placed in a line within the span of 1 inch, namely DPI (dots per inch). It

represents the spatial printing dot density and indicates the available printed

fine-feature size. Obviously, print resolution is greatly influenced by the

printed droplet volume. Smaller droplet size is able to provide a higher

resolution printing. Another ongoing measurement of the resolution of inkjet

printing is to use drop pitch in favor of metrication. Drop pitch gives the

spacing between the two adjacent printing dots, and has a direct relationship

with DPI as below:

(2.1)

For example, a resolution of 720 DPI equals to a drop pitch of about 35 µm,

indicating the inter-dot spacing of 35 µm. The resolution of drop-on-demand

inkjet printing is usually in the range of 70-100 µm. However, the droplet size

of commercial inkjet printers has been decreasing in order to achieve higher

print resolution. Droplets of 8-30 picoliters in volume, i.e., droplet size about

25-40 µm, can be generated by commercial inkjet printheads [36, 37].

In practice, due to the changes in droplet size and shape during the droplet

spreading and drying process, resolution of inkjet printing is also greatly

influenced by surface energy of the substrate and the conditions of the ink

solvent evaporation. By heating the substrate near or over the boiling point of

the ink solvent, the liquid in the droplet can be evaporated immediately upon


24 of 176

contact and providing a finer resolution than one in printed at room

temperature. However, it should be noted that by applying higher substrate

temperature also increases the risk of the nozzle clogging at the printhead.

Fundamental studies to obtain a constant width and shape of printed droplet,

lines and areas on the solid substrate are not fully developed.

2.1.3.2 Jetting Frequency

As a manufacturing tool in industry, product throughput or productivity for

inkjet printing systems is an important requirement, which is closely related to

the droplet jetting process and printing speed. The productivity of an inkjet

printhead is mainly determined by the jetting frequency, defined as the number

of droplets jetted from a nozzle within a certain time. As such, high jetting

frequency is desirable to attain high throughput for inkjet printing system.

On the other hand, print speed (the traveling distance in a unit time for a

printhead or motion stage) has the relationship with jetting frequency and

printing resolution (DPI), given as below:

(2.2)

This is shown that with the increase of the maximum jetting frequency, the

maximum print speed of an inkjet printer will be increased accordingly, and

thus extremely high throughput can be achieved. In 2009, Kyocera claimed the

world’s fastest inkjet printhead (KJ4B-JF06) with the jetting frequency up to

60 kHz [1]. With binary printing mode, each nozzle can eject 60,000 dots per

second (40,000 dots per second in multiple-value printing); with the 2,656

nozzles per printhead, the printhead can print approximately 150 million dots


25 of 176

per second. Its corresponding print speed can reach 150 m/min at the ultra

high resolution of 1200 × 1200 DPI.

Reliability is another important requirement for an inkjet printhead. For a

reliable jetting, a subsequent droplet should not be dispensed until the

meniscus oscillation from the previous droplet ejection cycle has sufficiently

damped [38, 39]. As this damping takes time, the maximum jetting frequency

is limited for the printhead. For a specific printhead, its maximum jetting

frequency is mainly dependent on the printhead construction and its driving

pulse signal [40]. Thus to obtain higher frequency jetting, a proper driving

waveform needs to be designed such that residual meniscus oscillation can be

effectively suppressed after each droplet ejection cycle. Typical drop-on-

demand inkjet printheads generate droplets at rates in the range of 0.1 to 10

kHz.

2.1.3.3 Drop Positioning Error

Inkjet printing process usually requires accurate placement of ink droplets into

predefined regions. Many of the error sources, affecting drop placement

accuracy, are applied in the design process to improve the stability and

precision of inkjet printing system [41]. Jet straightness and the location error

of the jet onto the substrate are two critical factors producing drop positioning

error. In Fig. 2.13, per nozzle straightness of the Spectra SX-128 inkjet

printhead is measured as shown for example. Jet straightness error less than ±5

mrads, i.e. ± 0.3°, shows a good performance in display applications [42].


26 of 176

Fig. 2.13: Jet straightness error in both X and Y directions for Spectra SX-128

printhead [42].

The total positioning error is a function of jet straightness, machine inaccuracy

and deviations in drop velocity, respectively [43]. Shimoda et al. [43] also

demonstrated that positioning errors derived from jet straightness and

mechanical inaccuracy can be compensated by a surface wetting effect caused

by a droplet and a substrate. It is worth pointing out that the flatness of the

precision substrate allows for small stand-off distances, which can reduce the

effect of flight-trace errors.

2.1.3.4 Nozzle Hydrophobicity Treatment

To improve printing performance of the printhead, the ink has to be jetted in

the form of complete droplets in a stable manner. Due to repeated droplet

ejection, the surface of a nozzle is wetted by the ink. As such, a nozzle surface

without proper hydrophobic treatment will be suffered from wetting. When

such wetting takes place, the droplet being jetted may be accumulated together

to form a lump on the nozzle surface, which will adversely affect the droplet

formation process. Both the size and placement accuracy of the ink droplets


27 of 176

ejected are then influenced. And finally, the quality of the printing is

deteriorated. Therefore, to guarantee a reliable inkjet printing, it is essential to

provide effective hydrophobicity treatment on the nozzle surface of an inkjet

printhead.

Fig. 2.14 shows total different inkjet printing behavior, for inkjet nozzles both

with and without hydrophobic treatment. It can be clearly seen that without

hydrophobic treatment, the jetted droplets are likely to accumulate on the

nozzle surface, which will deteriorate the printing or even totally stop the

printing.

Fig. 2.14: Two nozzles to show the effects of hydrophobic treatment. (a). Nozzle

without hydrophobic treatment. (b). Nozzle with hydrophobic treatment.

2.1.3.5 Inkjet-Printed Droplet Feature after Drying

Inkjet printing onto a nonabsorbent hard substrate is quite different from

printing onto absorbent paper due to the complex wetting and drying processes.

The shape or surface morphology of droplets and formed patterns are greatly

influenced by surface energy of the substrate and the ink solvent evaporation


28 of 176

process. The evaporation can be controlled by changing drying environment

such as substrate temperature, vapor pressure, etc. Generally, a circular droplet

on the solid surface evaporates from the boundary and flows outward. This

capillary flow tends to carry most of the dispersed materials to the boundary, a

phenomenon well-known as the coffee-stain effect [34]. It should be noted that

fundamental studies to obtain a constant width and shape of printed droplets,

lines and even areas on the solid substrate are not fully developed. In practice,

a considerable variety of the conditions of substrate surface and drop drying

can produce various dried droplet patterns. They are introduced in the

following sections.

Influence of Substrate Surface Wettability

Surface wettability of the substrate has a great influence on the printed droplet

feature after drying. Fig. 2.15 shows different shapes of printed droplets after

the ink solvent has evaporated, for both hydrophobic and hydrophilic substrate

surfaces [44]. On the contrary, the hydrophobic surface constrains the spread

of the droplet and gives a smaller drying droplet in diameter without the

coffee-stain effect.

Fig. 2.15: Image showing profiles of dried droplets printed on hydrophobic and

hydrophilic surfaces [44].


29 of 176

Influence of Drying Temperature

Fig. 2.16: Distinct dried droplet patterns under different temperature [45].

For super fine droplets such as ≤ 10 pl, the drying behavior of the droplets is

extremely sensitive to the substrate temperature. Under different range of

temperatures, printed droplet may achieve different dried patterns.

Representative images of dried droplet 3D profile and 2D cross-sectional


30 of 176

profile through their center are shown in Fig. 2.16 [45], wherein dried profile

of inkjet printed droplets under different substrate temperatures from 25 °C to

60 °C is characterized in order to obtain some reference information on

making a uniform thin film.

Dried droplet profiles that are achievable are categorized into three types:

Gaussian shape, transition shape and ring-like shape corresponding to low

temperatures (< 40°C), medium temperatures and high temperatures (> 50°C),

respectively [45]. At lower temperatures between 25 °C to 35 °C, droplets

finished their spreading stages before drying. Hence, the profiles obtained are

of Gaussian shapes. At temperatures between 40°C to 45°C, a change occurs

in shape of the dried profiles. This could be due to the higher evaporation rates

that prevented the droplet from spreading fully before it dried. At higher

temperatures above 50°C, droplets are dried immediately upon impact before

droplet spreading can occur. Therefore, droplets splash and simply solidify on

impact so as to form a ring-like structure.

2.1.3.6 Inkjet-Printed Line Morphology

Ideally, inkjet-printed straight lines would be smooth and even. But sometimes

it is difficult to fulfill all these features (straight, smooth and even)

concurrently. A few different behaviors come out when examining inkjet-

printed line formation under the various conditions such as droplet spacing,

droplet frequency, and substrate temperature. Fig. 2.17 shows five typical line


31 of 176

morphologies [46]. They can be labeled respectively as individual drops, a

scalloped line, a uniform line, a bulging line, and stacked coins.

Fig. 2.17: Examples of five typical inkjet-printed line morphologies. (a). Individual

droplets. (b). Scalloped line. (c). Uniform line. (d). Bulging line. (e). Stacked coins.

Droplet spacing decreases from left to right [46].

In the case of Fig. 2.17(c), Smith and Shin et al. [47] provided a calculation to

estimate the width of the printed line:

(2.3)

where d is the droplet diameter, w is the line width, θ is the contact angle, N is

the number of droplets printed for a line length L.


32 of 176

2.2 Squeeze Mode Piezo-Driven Printhead

As have been discussed in Section 2.1.1.2, there are four types of piezoelectric

drop-on-demand inkjet: squeeze mode, bend mode, push mode and shear

mode. This research is focused on squeeze mode piezo-driven DOD inkjet

printing.

2.2.1 Theory of Droplet Formation

2.2.1.1 Principle of Squeeze Mode Piezo-Driven Printhead

Fig. 2.18: Schematic representation of wave propagation and reflection in a squeeze-

mode piezoelectric inkjet printhead.

A typically squeeze mode piezo-driven DOD inkjet printhead is schematically

shown in the left top part of Fig. 2.18. The operating principle can be basically

explained as follows: when a voltage pulse is applied to the piezoelectric

actuator, it causes a sudden, quasi-adiabatic reduction of the ink chamber


33 of 176

volume, resulting in the generation of positive pressure waves inside the

chamber. These pressure waves propagate throughout and reflect inside the

chamber. Ink is propelled outwards when a positive pressure wave approaches

the nozzle. A droplet is ejected when the amount of kinetic energy transferred

outwards is larger than the viscous energy dissipation plus the energy needed

to form the surface of the droplet [48]. Besides the above two parts, the kinetic

energy left determines the initial velocity of the ejected droplet. For reliable

jetting, this initial velocity of a droplet needs to be several meters per second,

to overcome the drag action of ambient air [49].

The droplet formation process can be explained in more detail by referring to

the phenomena of wave propagation and deflection in the inkjet printhead,

which is a function of fluid properties, printhead design, and constituent

materials [50, 51]. The first investigation of such phenomena was conducted

by Bogy and Talke [52], followed by the extensive studies of other researchers

[53, 54]. As can be seen from Fig. 2.18, a typical bio-polar voltage signal

which normally used for printing is shown in the bottom of the diagram. In the

top part of the diagram is the propagation and reflection of the acoustic waves

related to each portion of that voltage signal.

At time “a”, the piezoelectric actuator moves radially outward due to a sudden

fall in the voltage. As a result, a negative pressure wave is produced inside the

ink. Point “b” represents that this pressure wave splits into two halves and

propagate in opposite directions to the reservoir side and nozzle side.

According to acoustic wave theory, the reservoir side can be treated as being


34 of 176

open since the inner diameter of the ink supply tube is comparable to that of

the ink chamber. Whereas the nozzle end can be treated as being closed since

the orifice diameter is negligible as compared to that of the ink chamber. The

pressure wave keeps its phase after reflecting from the closed end while

changes its phase to reverse after reflecting from the open end. The reason is:

for the open end, the boundary condition is zero pressure and this condition is

satisfied by superimposing an opposite phase pressure wave on the incident

pressure wave; the nozzle end is simply treated as a dead end with no flow rate

passing through it, to satisfy the governing equations for wave propagation,

pressure wave only changes its direction after reflecting from this end [50].

After travelling a distance of L, the two halves pressure waves travel back and

meet in the middle of the printhead chamber. At the same time “f”, an increase

of the voltage pulse causes the piezoelectric actuator to move radially inward,

which in turn producing a positive pressure wave propagating to the two ends.

The newly generated positive pressure waves coincide with the former two

halves reflected waves (“f” in Fig. 2.18) and superpose with them. As a result,

the negative pressure wave traveling to the open end is annihilated while the

positive pressure wave traveling to the closed end is doubled (“h” in Fig. 2.18).

This enhanced positive pressure wave propagates to the closed end and

reflects again from it at time , where c represents the speed of

sound in the ink. The drop ejection also occurs at this time “i”.

Bogy and Talke calculated the pressure history at the orifice and concluded

that the operation of DOD inkjet depends strongly on the length of the cavity.

They also found that the four important quantities in wave propagation were


35 of 176

linearly dependent on the cavity length and the sound speed in the ink. These

four quantities are: the optimum pulse width for generation of maximum

droplet velocity, which equals to ; the delay time for the first protrusion of

the meniscus, which equals to ; the period of meniscus oscillation,

which equals to ; and the period of low-frequency resonant and

antiresonant synchronous operation, which also equals to . The

calculations of these quantities have been extensively verified [40, 55].

The acoustic wave propagation theory is also utilized to help the printhead

design for high-speed inkjet. It is worthy to note that as a manufacturing tool,

high speed jetting is required to increase productivity of inkjet printing

technology. Thus there is a continuous requirement for increasing the

maximum jetting frequency for inkjet printing devices. However, for a reliable

jetting, a subsequent droplet should not be ejected until the pressure wave

from the previous pulse signal has sufficiently damped. This damping takes

time and thus limits the maximum jetting frequency [38, 56]. By utilizing the

self-sensing capability of the piezoelectric element, Kwon et al. proposed a

way to effectively measure the pressure wave inside the ink chamber through

the piezo current. Based on the measurement results, they designed a two-

pulse waveform which could greatly improve the jetting speed of the printhead

[40].

2.2.1.2 Droplet Generation Conditions

The governing equations for the 3-D drop-on-demand inkjet printing are:

0u v w

x y z

(2.4)


36 of 176

2

x x

p u v uu Vu

t x x x y x y

u wF g

z z x

(2.5)

2

y y

p v u vv Vv

t y x x y y y

v wF g

z z y

(2.6)

2 z z

p u w v ww Vw

t z x z x y z y

wF g

z z

(2.7)

where is the velocity vector with u, v and w in the x-, y- and z-axes; p, ρ and

µ represent the pressure, density and dynamic viscosity of liquid, respectively;

and denote the gravitational force and the surface tension force at the

liquid-gas interface [57]. There are hydrostatic pressure, viscosity force, fluid

inertia and surface tension force which will dominantly influence liquid flow

in a drop-on-demand device. It can be seen from Fig. 2.19 that to eject a

droplet out of the nozzle, the acoustic energy obtained from the piezoelectric

actuator must be sufficient to compensate the energy loss due to viscous

dissipation and the formation of liquid free surfaces.

In a fluid system with characteristic velocity V and characteristic length L, the

Weber number, ，is a measure of the relative importance of

the fluid’s inertia compared to its surface tension. Wang et al. [19] claimed

that it was reasonable to choose as the condition for droplet

generation, giving a initial droplet velocity . For simplicity,

the acoustic pressure was treated as a square hydrostatic pressure pulse with a


37 of 176

magnitude of P0 and a pulse width of t0. According to the law of conservation

of energy, the kinetic energy of the ejected droplet with such a velocity V,

should equals to the work done by the acoustic pressure. As a result, P0 can be

estimated. Later this hydrostatic pressure was refined by adopting a

commercial inkjet simulation tool: Conventor Inkjet Developer. Table 2.1

shows the minimum hydrostatic pressure required to generate different size of

water droplets. The results indicate that the required actuation pressure

increase is almost directly inversely proportional to the droplet size, which is

reasonable as surface tension is the dominant factor influencing droplet

generation [58].

Fig. 2.19: Schematic representation the basic energy requirement for ejecting a

droplet.

Table 2-1: The minimum actuation pressure for droplet generation in DOD inkjet

devices [58].

d (μm) 10 2 0.4 0.1

Volume (pl) 7.1e-1

6e-3

3.8e-5

7.1e-7

P0 (MPa) 0.2 1 4 20

Besides nozzle size, the fluid properties of the ink also influence droplet

formation. Fromm rescaled the governing equations (eq. 2.4 to eq. 2.7) and


38 of 176

used the following grouping of fluid properties to provide a dimensionless

analysis of the mechanics of drop formation in DOD inkjet [49]:

(2.8)

The relation is equivalent to the inverse of the Ohnesorge number Oh, where ,

ρ, and are the viscosity, density, and surface tension of the liquid,

respectively, and d is the characteristic length. For an inkjet printhead, d is the

diameter of the nozzle orifice. Fromm predicted that successful droplet

ejection was only possible when Z > 2 and that for a given pressure pulse

droplet volume increases as the value of Z increase [49]. This prediction was

later refined by Reis and Derby, who carried out experiments for dispensing

different concentrated alumina wax suspensions and then concluded that

successful DOD printing could be achieved in the range 1 < Z < 10 [51]. Here

the lower limit represents no droplet ejection due to too much energy

dissipated by viscosity forces, and the upper limit represents the formation of

satellite droplets. In reality, inkjet printing can be reliably utilized for

industrial applications even when satellite droplets are produced, just provided

that the satellites can merge with the main droplet. For example, polymer

solutions with Z numbers bigger than 50 have been successfully dispensed by

Schubert et al. [59].


39 of 176

2.2.1.3 Droplet Velocity and Droplet Size

Fig. 2.20: Effects of pulse amplitude on droplet velocity and droplet volume [60].

The driving signal to the piezoelectric actuator has significant influence on

droplet formation. Results from different research groups have shown that the

droplet velocity and droplet volume are linearly dependent on the amplitude of

the driving pulse (for simple rectangular input pulse), as shown in Fig. 2.20.

This is understandable as referring to the fractional volume change due to the

piezoelectric effect:

(2.9)

where V is the volume of the piezoelectric actuator, d31 is the piezoelectric

strain constant, U is the applied voltage and t is the thickness of the

piezoelectric tube [61]. The negative sign indicates contraction when the

applied pulse has the same polarity as the original polarizing voltage for the

piezoelectric element. From eq. 2.9 is can be seen that the volume change of

the piezoelectric tube is liner to the amplitude of the input pulse. An increase

in voltage amplitude will lead to a greater volume change in the ink chamber


40 of 176

and correspondingly larger induced acoustic waves and fluid acceleration,

ultimately, a bigger droplet with higher velocity.

Fig. 2.21: Effects of pulse width on droplet velocity and droplet volume [60].

Fig. 2.21 shows the effects of pulse width on droplet velocity and droplet

volume. Both of the quantities exhibit a maximum as the pulse width is varied.

Derby et al. showed that the location of this maximum value of droplet

velocity remained unchanged when the driving pulse amplitude was increased,

but it shifted when the fluid properties were changed [62]. Similar behavior

was also observed for droplet volume. Derby et al. also observed the periodic

behavior of droplet as the driving frequency f of the piezo signal is varied, as

shown in Fig. 2.22. The x-axis represents “inverse of the jetting frequency”,

rather than the “pulse width”. They are two different parameters, although

both of them have a unit of micro-second. In practice, pulse width is normally

smaller than the “inverse of the jetting frequency” as there are intervals

between successive piezo signals. From Fig. 2.22, this periodicity

phenomenon is also dependent on the fluid properties. This is reasonable as

the maximum in droplet velocity correspond to conditions of resonance, which


41 of 176

are dependent on the ink chamber geometry as well as the speed of acoustic

wave in the ink, which is dependent on the ink properties.

Fig. 2.22: Effects of jetting frequency on droplet velocity and droplet volume [62].

2.2.1.4 Satellite Droplet

As have been mentioned before, Reis and Derby predicted that satellite

droplets would be produced during inkjet printing when the Ohnesorge

number smaller than 0.1 [51]. Since satellite droplets can negatively impact

the printing resolution, there has being a great interest in understanding the

satellite droplet formation process, and also the methods to avoid it [63-65].

A sequence of images captured during DOD droplet formation is shown in Fig.

2.23, revealing the main features of this process. When the piezoelectric

transducer contracts, ink inside the chamber is accelerated and squeezed out of

the nozzle. Initially, the meniscus extends rapidly outward until a liquid


42 of 176

column with a round leading edge is formed (images 1-3 in Fig. 2.23). After a

short time (normally several tens of micro-seconds after the contraction

behavior of the transducer, starting at approximately 12 µs in the figure, as

image 4), the outward flow rate decreases. The resulting difference in Z-axial

velocity between the column head and the liquid at the nozzle exit causes the

liquid column to stretch. The velocity of the liquid at the orifice continuously

decreases until no flow into the column, or even decreases further due to some

inverse flow caused by the extension of the piezoelectric transducer to its

equilibrium position. The liquid column is continuously extended due to the

inertia and this extension rate decreases with time as new surface is formed

which resulting the increase of surface energy. With the extension and

stretching of the liquid column, the liquid at nozzle exit necks, with a necking

position at the nozzle exit. Almost concurrent with the liquid column

stretching, a second necking point appears (image 5 in Fig. 2.23) at the top of

the column head. Finally, the liquid thread pinches off from the nozzle,

forming a free long liquid thread with a bulbous head [63]. Following the

pinch-off, the tail recoils toward the bulbous head, to reduce the surface

energy of the whole liquid column. As has been mentioned in above, there is a

new neck near the bulbous head. The new necking continues to evolve until

the liquid thread pinches off from the bulbous head, forming a primary droplet

and a new free thin liquid thread. The lower end of this liquid thread recoils

rapidly toward the upper end, again, to reduce the surface energy of the thread.

Depending on its length, this new liquid thread may shrink into a smaller

droplet (or satellite, image 16 in Fig. 2.23), or break up into two or even more

droplets [63].


43 of 176

Fig. 2.23: Sequence of images of DOD droplet formation for water [63].

To conclude, the main reason for satellite generation is due to the existence

and later break up of a long liquid thread which connecting the bulbous head

with the nozzle exit, before the liquid column pinching off from the nozzle.

Dong et al. [63] predicted that this liquid thread would contract and combine

into the bulbous head without breaking up if the length of liquid thread at

pinch-off, lb, does not exceed a limiting value .

(2.10)

(2.11)

(2.12)

where is the nozzle radius, is a constant for water, is the first pinch-

off time, is the second pinch-off time, is the ejection time (time from

emergence of liquid form the nozzle until the ejected volume reaches its

maximum), is the capillary time. is the grow rate of the most

unstable disturbances. To suppress satellite droplets formation, a bigger is

desired. According to Dong et al., this can be achieved by increasing liquid

viscosity and decreasing surface tension, or optimize the piezoelectric

waveform. Another common attempt to eliminate satellite droplets generation

is to reduce the magnitude of the piezoelectric pulse, which will decrease the


44 of 176

droplet ejection velocity and may shorten the liquid thread. However, droplets

with low jetting velocity suffer from the droplet scattering and the loss of

accuracy in positioning [66].

As can be imaged, the formation of the liquid thread connecting the primary

droplet and the nozzle exit is related to the physical properties of the liquid

itself. Shore et al. [65] found that by adding small amounts of polymers in

Newtonian solvents, satellite droplets generation were effectively eliminated.

This is understandable as when the polymer molecules are stretched out during

the droplet formation, the elasticity of the solution will cause the liquid thread

which connecting the primary droplet and the nozzle to snap back and

combine with the primary droplet, forming a monodisperse droplet. While

remember that this liquid thread normally pinches off from the primary droplet

and breaks up into satellite droplets for a purely Newtonian fluid. Shore et al.

also found that to efficiently suppress satellite droplets formation, there was a

minimum required concentration for the polymers to be added, for a fixed

molecular weight. Consequently, the resultant elasticity of the solution will

increase and a stronger piezoelectric pulse is required to eject the droplet. The

solution containing polymers will also have a longer liquid thread, a longer

droplet separation time, and a lower droplet velocity as compared to the purely

Newtonian fluids with similar shear viscosity [65]. Furthermore, the method

requires the change of the ink properties, which might be undesirable for

practical applications.


45 of 176

2.2.2 Printhead Fabrication

2.2.2.1 The Overall Printhead Structure

Fig. 2.24: Different kinds of commercial printheads.

Currently, printheads for research purposes can be bought from companies

such as Microdrop and Microfab. Fig. 2.24 shows several dispensers from

these companies. Regardless of the internal components it contains, the basic

design idea for all the squeeze mode piezo-driven printhead is the same. Fig.

2.25 schematically shows the construction of a typical piezoelectric squeeze

mode printhead. By using epoxy adhesive, a piezoelectric element is tightly

attached onto a conventionally used glass tube which with an orifice at one

end. When an electrical pulse is applied, the piezoelectric element will

contract inward, squeezing the glass tube as well as the liquid inside, and

ejecting a droplet from the nozzle. The piezo tube is made from a ceramic

material that changes shape when a voltage is applied to it. There are several


46 of 176

kinds of piezoelectric ceramic materials: PZT-5H, PZT-5A, PZT-8 and PZT-4.

Different materials have different piezoelectric constants thus have different

stresses-strains relationships [67].

Fig. 2.25: Schematic of the construction of a piezoelectric squeeze mode DOD

printhead.

A new printhead with interchangeable nozzles will be designed in this study.

Accordingly, the printhead will be separated into two main parts: the printhead

chamber and the interchangeable nozzle.

2.2.2.2 Ejection Nozzle Requirements

Lee [68] claimed that an ideal ejection nozzle should have a tapered cross

section ending up in a short hole with an aspect ratio round one-to-one. The

purpose of the overall conical taper is to minimize the flow resistance without

compromising mechanical strength. To illustrate the effects of different aspect

ratios on nozzle behavior, three types of nozzles, a tapered one with a

proposed aspect ratio, a very thin one and a very thick one, are compared. As

shown in Fig. 2.26, the left one in a proposed geometry works properly during

the droplet generation cycle. However, air can be easily sucked into the ink

chamber for the second nozzle, during the fluid withdrawal phase. The

existence of air bubbles will cause an increase in the ejection threshold due to

air bubbles absorbing the pressure impulse which is supposed to squeeze out


47 of 176

the ink. Worst case is the stop of jetting due to big air bubbles inside the ink

chamber. Air entrapment will not occur for the nozzle has a high aspect ratio,

as the third nozzle shown in the figure. However, the local flow resistance is

too much for this nozzle and thus will also raise the threshold for jetting. To

conclude, a desirable nozzle should have an orifice shape similar to that of the

first one.

Fig. 2.26: Ejection nozzle orifice cross section requirements.

2.2.2.3 Ejection Nozzle Fabrication Methods

Various techniques can be used to fabricate small holes. A definitive paper on

such techniques is published by [69].

Thermally Tapered Glass Pipettes:

By simply heating a glass tube and pulling it while hot, a closed cone with an

included angle will come out. This closed end is then polished until a hole of a

desired diameter is produced. All the nozzles of the printheads shown in Fig.


48 of 176

2.24 are fabricated by this method. It is a simple and low-cost fabrication

method. Meanwhile, it has following disadvantages. Firstly, it is difficult to

achieve a perfectly axially symmetric nozzle, due to the polishing of a conical

tip. This asymmetric nozzle will produce offset jets, which will destroy the

accuracy of printing, especially when the jetting velocity is relatively low.

Furthermore, glass material is so fragile that a simple touching or wiping the

nozzle tip against a rigid object may destroy the orifice and, again, lead to an

asymmetric nozzle. Secondly, the fabrication method is so sensitive to the

heating process that it is virtual impossible to replicate the exact profile of the

nozzle. This nonuniformity is obviously undesirable for industrial applications

that require parallel, identically operating channels of interchangeable spare

units [68]. For these reasons, a silicon micromachining method is proposed for

fabricating inkjet nozzles.

Silicon Micro-Machined Nozzles:

For crystal wafers, slice orientation is used to denote the crystallographic

orientation of their surface. It is well-known that there are three most common

used slice orientations for silicon wafers: (100), (110) and (111). Accordingly,

the surfaces perpendicular to such orientations are denoted by (100) plane,

(110) plane and (111) plane, respectively. For a (100) silicon wafer, such

orientations and planes are shown in Fig. 2.27. Wet orientation-dependent

etches (ODEs) can be used to etch silicon wafers for nozzle fabrication. It is

an anisotropic etching method at rates that vary with crystallographic direction

[70]. For example, for KOH etching at 85 ˚C,


49 of 176

where , and represents the ODEs’ etching rate at (111),

(100) and (110) plane, respectively. Since is negligible as compared to

and , for (100) a silicon wafer, the etching process will proceed

downward until (111) planes are reached, forming a V shape, as shown in Fig.

2.27(b) and Fig. 2.27(c).

(a)

(b) (c)

Fig. 2.27: KOH etching for a (100) silicon wafer. (a). Slice orientations for silicon

material. (b). Slice orientations shown in a plan view of a (100) silicon wafer. Etching

process proceeds downward until (111) planes are reached. (c). “A-A” cross-section

view.

To obtain an inkjet orifice with a desired diameter, a silicon wafer is firstly

etched by chemical anisotropic (KOH etching is normally used) method to

form a deep pyramidal pit, just leaving a short distance to the opposite end of

the wafer. Then Deep Reactive Ion Etching (DRIE) is used to form a


50 of 176

cylindrical aperture of a desired diameter through this thin layer, as shown in

Fig. 2.28.

(a)

(b)

Fig. 2.28: Nozzle fabricated by silicon micromachining method comprising KOH

etching and Deep Reactive Ion Etching. (a). Plan view of the etched wafer. (b). “A-

A” cross-section view of the etched wafer.

The optimal thickness of this thin plateau, or the optimal thickness of the

aperture, is determined by the diameter of the orifice. An aperture thickness of

one to two orifice diameters is sufficient to suppress air entrapment. Another

concern is the thickness of this plateau should be enough to ensure sufficient

mechanical strength, avoiding nozzle damage due to the common nozzle

cleaning process such as being immersed in an ultrasonic bath, or the

accidental touching of the nozzle against a rigid object [68].


51 of 176

Silicon micromachining method is more complicated and more costly as

compared to the glass heating and pulling method; however, the method

provides good mass productability and exact reproducible desired nozzle

profile for large production runs [71-73].

Electro-Discharge-Machining (EDM)

In EDM method, a spark discharge between a cutting electrode and the

workpiece is utilized to remove material. EDM can cut extremely hard

material to very close tolerance, but it also has the disadvantages of inability to

cut non-conductive materials and a slow cutting rate.

Laser Drilling

To date, laser drilling can produce holes as small as 1 micron. Due to its non-

contact feature, it can drill holes on curved surfaces and can handle both hard

and soft material.

Plenty of methods have been adopted to fabricate nozzles for inkjet purpose.

This research is more concentrated on the fabrication of glass nozzles.


52 of 176

2.3 Creation of Ultra-Small Droplets

2.3.1 Needs for Generation of Ultra-Small Droplets

The manufacture of flat panel displays (FPD) for computer monitors and

televisions are a $ 60 billion industry. The state-of-the-art facilities are capable

of fabricating panels on ~ 2 m × 2 m substrate, and the substrate size has

doubled every two to three years since 1990 [44]. Traditional display

manufacturing utilizes photolithography techniques, which is a well-

established patterning method for silicon integrated circuits (ICs). However, to

build the photolithography and etching systems for huge substrates is

challenging and extremely expensive. From this viewpoint, emerging direct

printing of functional electronic materials [74-77] has attracted consideration

attention. Inkjet printing is a data-driven direct-write method without the

complicated photolithography and etching processes, thus leading to a great

reduction in manufacturing cost and processing time.

However, the requirements for patterning display pixel are more challenging

than the traditional document printing: a document pixel element is formed by

a drop of ink, while a display pixel is a circuit comprising different materials

[44]. Currently, there is a wide gap between the resolution required for display

manufacturing and the resolution of a typical inkjet device. Consequently, a

more precise ink jet system that capable of producing subpicoliter or even

subfemtoliter droplet is required. It is also the most critical requirement for

applying inkjet printing to the fabrication of high-performance electronics


53 of 176

devices, in where thin-film transistors (TFTs) with a channel length of 1 to 2

µm is commonly required [78, 79].

Fig. 2.29: Schematic of photolithographically predefined inkjet printing. (a).

Schematic diagram of high-resolution inkjet printing onto a prepatterned substrate.

(b). AFM showing accurate alignment of inkjet-printed PEDOT/PSS source and drain

electrodes separated by a repelling polyimide (PI) line with L = 5 µm. [20]

Actually, several methods have been proposed and verified to eliminate such a

gap between the resolution required for those high-performance electronic

devices and the typical resolution of conventional inkjet printing devices. For

example, photolithographically predefined features [18, 20, 80, 81] or surface

pretreatment [15, 82] in the form of hydrophobic and hydrophilic patterns can

confine and guide the flow and spread of the printed droplets when they land

on the substrate. As shown in Fig. 2.29(a), a photolithographically predefined

substrate was fabricated as follows: a glass substrate with a 500 Å

hydrophobic polyimide film was spin-coated with photoresist. The photoresist

above the source-drain regions on the bare glass was removed and the

corresponding parts of polyimide were etched through. Then Oxygen plasma

was used to burn up the hydrophobic groups of the source-drain regions and

make it hydrophilic, whereas the polyimide line (PI line in Fig. 2.29(a))

defining the thin film transistor channel was still covered and protected by

photoresist and remained hydrophobic [20]. Lines of PEDOT droplets were


54 of 176

deposited into the source-drain regions at a distance d from the mid polyimide

line. This distance d was small enough to ensure that the spreading droplets

could reach the repelling line. Atomic force microscopy (AFM) showed that

the deposited PEDOT electrodes extended accurately up to the repelling

polyimide line, without destroying the narrow gap (5 µm) between them.

With above methods [15, 18, 20, 80], the gaps between printed droplets can be

controlled at the submicrometre level, which is important for electronics

fabrication as such gaps define the transistor channel lengths. However, above

methods do not offer a universal approach to high resolution. Furthermore,

complicated photolithography and etching processes are required to deal with

the substrate. By this token, reducing droplet size maybe the most general and

direct way to obtain high printing resolution. Actually, inkjet printing

technology is also developing at a rapid pace and the size of the dispensed

droplets halved every four-and-half years during the last two decades, in a way

similar to Moore’s Law for transistors [83]. Thus, it can be foreseen that a

more precise inkjet system that is capable of producing ultra-small droplet will

be designed and fabricated in the near future.


55 of 176

2.3.2 Methods for Printing Ultra-Small Droplets

2.3.2.1 Reducing Nozzle Size

Typically drop-on-demand systems eject droplets with the radius roughly the

same as the radius of the nozzles. To date the most direct and reliable way of

reducing the droplet volume has been reducing the orifice size. Wang et al.

[19, 58] have demonstrated the fabrication of inkjet nozzles with orifice size

as small as 2.5 µm, based on silicon micro-machining technology. Their

printhead was composed of a large array of inkjet devices, operated in the

thermal bubble mode. Stable generation of water droplets down to 3.5 µm has

been demonstrated. Kung et al. [84] have reported the generation of 3 to 4 µm

water droplets by a piezo-driven printhead with a very small orifice measuring

only 1 µm in diameter. Their nozzle tips were fabricated by heating and

pulling 3-mm-diameter Pyrex tubing. The diameter of the dispensed droplets

reduced from 4 to 1 µm after a short travel distance (several millimeters), due

to evaporation. Although micro-scale droplets have been successfully

generated by reducing the size of the printhead orifice, the problem of

clogging and breaking of the nozzle becomes an obstacle to reliable operation.

Thus there is a need for effectively reducing droplet volume without reducing

nozzle radius in DOD inkjet printing.

2.3.2.2 Controlling of Waveform

The phenomenon of much smaller, faster droplets ejected from an inkjet

printhead with a relatively bigger orifice was firstly described by researchers

from Xaar. The mechanism was explained by Temple [85], who modeled the


56 of 176

acoustic waves of the fluid in the printhead chamber and claimed that a special

configuration of interacting pressure pulses is responsible for the phenomenon.

Fig. 2.30: Schematic of pulse waveforms used for driving the inkjet printhead. (a). A

uni-polar waveform. (b). A bi-polar waveform. (c). The new waveform for small

droplet generation. [21]

Fig. 2.31: (a) – (c) Images showing appearance and disappearance of a tongue and

formation of droplet with a diameter similar to that of the nozzle. (d) – (f) Images

showing formation of a droplet with a diameter much small than that of the nozzle

orifice. [21]


57 of 176

Later Basaran et al. [21] confirmed that droplet with one order of magnitude

reduction in its volume could be produced by judiciously controlling the

capillary, viscous, and inertial time scales that govern the flow within the

nozzle and the forming droplet. They investigated the effects of different piezo

driving waveforms on droplet formation. Their results show that as compared

to a conventionally used uni-polar signal (Fig. 2.30(a)), a bi-polar piezo (Fig.

2.30(b)) signal will cause a “tongue” protruding from the tip of the primary

droplet, as shown in Fig. 2.31(a). The tongue is resulted from a small volume

of liquid traveling at a high velocity relative to the surrounding liquid.

However, the tongue was found to be absorbed into the primary droplet later,

forming a droplet with a similar diameter to the nozzle orifice, as shown in Fig.

2.31(b). In order to suppress the formation of the large primary droplet and

help the tongue detaching from the primary droplet, a new waveform (Fig.

2.30(c)) consisting of a succession of three square-wave pulses is suggested.

As shown in Fig. 2.31(e), the tongue successfully detached from the primary

droplet, forming a much smaller droplet.

Similar effects of reducing droplet size are also found when utilizing M-

shaped, W-shaped, and other types of waveforms [23, 24, 25]. However, all

above methods only work over a limited range of Ohnesorge numbers [21] and

their effects are also limited: the diameter of the dispensed droplets can be

only reduced to a maximum of 60 % of the orifice diameter. Different methods

are required to reduce droplet size to micro-scale or even submicro scale.


58 of 176

2.3.2.3 Electrohydrodynamic Jetting

Fig. 2.32: Schematic of an electrohydrodynamic jet system. [86]

Electrohydrodynamic jet (E-jet) process is a unique and versatile jet-based

technique, which utilizes electric field, rather than acoustic or thermal energy,

to create the fluid flow required for ejection of liquid [86, 87, 88]. Fig. 2.32

shows a schematic diagram of a typical e-jet system. It comprises of a liquid

reservoir, a pneumatic pressure controller connected to a needle, and a

grounded electrode located centrally below the needle. By applying a potential

difference between the needle and the grounded electrode, the liquid inside the

needle can be charged. This charged liquid exits the needle and enters the

high-intensity electric field, forming different liquid geometries from which a

jet or multi-jets, or even dispersed droplets evolve. The size of these jets or

droplets can be controlled by the intensity of the electric field, the liquid

properties, and the flow rate into the needle. As has been mentioned before,

conventional inkjet devices generate drops with diameter approximately the

same as that of the nozzle orifice. In contrast, e-jet does not suffer from this


59 of 176

disadvantage and droplets of a few micrometers in diameter can be produced

from needles that are a few hundred micrometers in diameter.

To better understand the fundamental dynamics of this electric-field-driven

jetting behavior, sequential images of the liquid ejection process were

obtained by reference Marginean et al. [89] with a high-speed-video camera.

As shown in Fig. 2.33, the meniscus at the needle tip expands and contracts

periodically due to the electric field. Correspondingly, a complete jetting cycle

is divided into four stages: liquid accumulation, formation of the well-known

Taylor cone [90, 91], droplet ejection (or jet ejection), and relaxation. The

entire sequence takes about 0.5 ms.

During the first stage, the liquid accumulates at the end of the needle tip due to

the net flow from the liquid reservoir. The almost spherical meniscus indicates

that the surface tension is the dominant force during this stage. The application

of the electric field will cause mobile ions in the liquid to accumulate near the

surface of the liquid meniscus. Consequently, a tangential electrostatic stress,

known as the Maxwell stress [86], will be induced on the liquid surface, due to

the mutual coulombic repulsion between these ions. With the accumulation of

surface charges, the initial spherical meniscus changes its shape gradually into

a conical form under the tangential Maxwell stress, as shown in the second

stage. The decrease in the radius of curvature at the cone apex continues until

the Maxwell stress matches the maximum capillary stress. At sufficiently high

surface charges, the Maxwell stress overcomes the capillary tension and jet (or

droplets) ejects from the apex (the third stage, 252-412 µs) to expel some of


60 of 176

these surface charges (known as the Rayleigh limit). Both the cone volume

and the charges decrease due to the ejection, resulting in a Maxwell stress less

than the capillary tension. As a result, the ejection stops; a fast retraction of the

liquid is observed and the cycle repeated, as shown in the fourth stage.

Fig. 2.33: Time-lapse images of the pulsating Taylor cone with the four stages of the

complete jetting cycle. Each frame is an average of 100 exposures with the same

delay. [89]

The above jetting process is well-known as the pulsating mode jetting. It is

found that at sufficiently high electric field, a stable jet mode could be

obtained, in which a continuous liquid stream rather than pulsating jet

emerging from the needle. Both jetting modes can be used for high-resolution

printing, while the pulsating mode jetting might be preferred in the sense of

jet-on-demand, as in drop-on-demand inkjet printing. Different research

groups [78, 86, 92, 93] have demonstrated the success of the method in


61 of 176

patterning structures with critical dimensions as small as 1 µm, as shown in

Fig. 2.34.

Fig. 2.34: High-resolution e-jet printing with printed feature size smaller than 1 µm.

[86]

It is worth noting that the complete jetting cycle for above pulsating mode e-

jetting normally lasts for 3 to 10 ms [86], which corresponding to a maximum

jetting frequency of around 300 Hz. Unlike using DC high electric field in

pulsating jetting, there is another type of e-jet, the pulsed-voltage mode, in

which pulse waves of high voltage are used to switch electrohydrodynamic

force (“on” or “off”) [94, 95-98]. It also requires more than 3 ms [95] for

Taylor cone formation, after applying the pulsed voltage. Consequently, it is

also theoretically impossible to produce jets (or droplets) of a frequency

higher than 300 Hz. Kim et al. [99] demonstrated that this maximum jetting


62 of 176

frequency could be increased to 5 kHz, by combining electrohydrodynamic

force and the mechanical actuation from a traditional drop-on-demand

printhead. The basic idea of this so-called hybrid jetting system (HJS) is: by

utilizing the expansion and contraction of the piezoelectric element, to

accelerate the “liquid accumulation”, “Taylor cone formation” and the

“relaxation” stages during the pulsating mode e-jetting. As a result, the jetting

frequency can be increased [100]. However, this hybrid jetting technique is

still of low throughput, as compared to the conventional drop-on-demand

inkjet which generally has a maximum jetting frequency above 10 kHz.

To conclude, the e-jetting technique allows the generation of ultra-fine

droplets (or jets) down to 1 µm; however, at the cost of reduced throughput.

Furthermore, it has rigid restrictions on the liquid to be used, such as the

conductivity and other physical properties.

2.4 Organ Printing - Science Rather Than

Fiction

Positioning of living cells in a desired pattern onto a substrate is extremely

important to cell-based technologies, including the fundamental investigation

of cell functions and tissue engineering [101, 102]. The more exciting thing is

that nowadays the self-organizing properties of cells and tissues are used by

material scientists and tissue engineer to build 3-D living structures, as shown

in Fig. 2.35. The generation of such structures forms the basis of tissue

regeneration and possibly, the fabrication of implantable organs [2].


63 of 176

Fig. 2.35: Printed cells. (a). 3-D tube structure made from printed cells. The image

shows an inner layer of human umbilical endothelial cells (green) and an outer layer

of human aortic smooth muscle cells (red). (b). Printed yeast patterns after 3 days of

culture. [2]

2.4.1 How to Realize

Inkjet printing is a suitable candidate for organ printing. The power of inkjet

printing lies in its ability to deliver picoliter volumes of materials at high

speed and accuracy on a target interface (probably non-planar surface), and to

deliver active substances to a developing structure in a well-defined timing

sequence. By using different cell types as different bio-inks, and delivering

them to exact positions to mimic tissue structures of the original tissue, inkjet

printing offers a possible solution for building whole structures such as bone,

cornea, ligament, cartilage etc, to solve the organ transplantation crisis. It has

been used successfully in a similar manner for automated rapid prototyping

technology which precisely positions droplets onto a substrate. However,

printing living cells into a desired structure which can ultimately grow to an

implantable organ is a much more challenging task.


64 of 176

Firstly, cells have to survival from the shear stress during the printing, and

keep their viability. Fortunately, to date, many different cell types have been

printed successfully by different printing methods and their viability has been

verified [3-10, 103]. By dispensing human fibroblast cells through a 60 µm

nozzle, Saunders et al. carried out a comprehensive study to investigate the

relationship between cell survivability and the inkjet printing parameters [9].

Their study supported previous claims [4, 5, 8, 10] that cell survivability was

not significantly affected by the printing process since cell survival rates only

fell from 98% to 94% when the excitation pulse was increased from 40 to 80

V. However, in all above studies, the printing process was carried out with

relatively bigger nozzle diameters (normally bigger than 60 µm) and lower

droplet velocities (1.0 to 3.0 m/s). These limitations may be of importance,

because the shear stresses, which are expected to be the main factor in the

killing of cells during the printing process, are proportional to the velocity

gradients within the nozzle. In fact, shear stresses have been studied

extensively to predict the damage of animal cells suspended in various laminar

or turbulent flows [104-107]. To conclude, high rate of cell death might be

possible during printing when smaller nozzle and high droplet velocity are

required.

After verifying viability of the printed cells, the second step towards organ

printing should be successful generating of 2D cell patterns. The number of

cells inside each printed droplet will be one important factor for reliable cell

printing, as “empty droplet” and cell-less droplets may be undesirable. For

well prepared cell ink, this number mainly depended on the cell concentration

of the cell ink as well as droplet volume [8, 103]. However, during the printing,


65 of 176

cell agglomeration or sedimentation may be occurs [9, 108], thus undesired

variations in this number can be produced. The most direct and effective way

to avoid such “empty droplet” should be increasing cell concentration of the

cell ink. Another important issue is that the printed cells should be able to

adhere, spread and proliferate on the substrate. Thus cells are normally printed

on gels, such as alginate or collagen. Alginate has been increasingly utilized in

tissue engineering to support encapsulated cells and to regulate cells function,

in a manner similar to the extracellular matrices of mammalian tissues [109].

Alginate’s popularity comes from its advantages of biocompatibility,

nonimmunogenicity [110] and gentle gelling behavior [111]. However，the

major limitation to its use as extracellular matrices is that alginate does not

mediate mammalian cell adhesion [112, 113]. To promote cell adhesion within

alginate gel, ligands such as arginine-glycine-aspartic acid (RGD) [114-117],

GRGDY [118], KGD and VAPG [115] have been used. Collagen is another

widely used hydrogel with a number of advantages including biodegradability,

low immunogenicity and controllable stability. Furthermore, collagen contains

cell adhesion domain sequences such as RGD, thus can facilitate cell adhesion

for anchorage-dependent cell types [119, 120].

Thirdly, the most ambitious, also the most challenging step is to create 3D

living structures. The conversion from cell suspension into 3D organ

structures needs to be guided by 3D scaffolds as cells normally do not self-

assemble into organ-like structures [122]. There are two possible ways to

incorporate cells into such a scaffold. The first method is to create a

degradable 3D scaffold first, and then seed cells into it, as shown in Fig. 2.36.


66 of 176

The scaffold has following important functions: serving as an adhesion

substrate for the cell, promoting cell proliferation and cell-specific matrix

production [123]; providing mechanical support in the initial tissue growth

stage; guiding the development of new tissues with appropriate function [124].

The porosity and the internal pore organization of the scaffold have an

important influence on its biodegradation dynamics, mechanical stability and

nutrient diffusion, as well as on cell migration [125]. Conventional methods

for fabricating scaffolds include phase separation [126], particulate leaching

[127], gas foaming [128], freeze-drying [129, 130] and electrospinning [131].

Fig. 2.36: 3D scaffold and the cells seeded into it. (a). A 3D scaffold fabricated by

rapid prototyping method. (b). Big view of the scaffold shown in (a). (c). Human

fibroblast cells seeded into a 3D scaffold, after 18 days of culture. [121]

Fig. 2.37: Fabrication of a scaffold by 3D plotting. (a). One layer. (b). Two layers.

[122]


67 of 176

However, above processing methods have inherent limitations in precise

control of pore size, pore geometry, pore interconnectivity, spatial distribution

of pores, and construction of internal channels within the scaffold [123].

Different new techniques have been proposed to eliminate these limitations,

among them, a new Rapid Prototyping technology based upon 3D plotting

technology, was developed to produce scaffold with complex architectures

according to computer design [132]. A key feature of the technology is its

ability to create 3D structures from liquids and pastes in liquid media. As

shown in Fig. 2.37, by positioning individual microdroplets and in situ

bonding them, it is possible to layer-by-layer fabricate scaffolds with desired

interconnecting pore design, thus meeting the demands for cell attachment and

cell growth. Currently this method has been widely used for 3D scaffold

manufacture [123, 133, 134, 135].

It is worth noting that above method of creating 3D living structures is based

on the premise that seeding cells into porous biodegradable scaffolds will be

sufficient to generate organs. However, Boland et al. [136] claimed that there

were at least four limitations for the method:

cell penetration and seeding is still far from optimal;

organs generally consist of several different cell types, it is a

challenging task to “seed” different types of cells in specific positions

of the 3D scaffold;

the rigid, solid scaffolds are not optimal for engineering contractile

tissue, such as heart and vascular tubes;


68 of 176

absence of vascularization is the main problem with using solid

scaffold seeding technology.

Consequently, there is an increasing interest in the using of the second method,

in which the 3D living structures is fabricated by directly incorporating cells

into the scaffold fabrication process. As shown in Fig. 2.38, cells and a kind of

“thermo-reversible” gel are positioned by inkjet printheads, to form alternate

layers onto a glass slide. This method is also termed as “organ printing” [137].

The closely packed layers will coalesce providing that the alternating layers

are thin enough. After the whole tissue grows up, the gel will degrade by

simply cooling it.

Fig. 2.38: Schematic diagram of organ printing. [138]

Herein the “thermo-reversible” gel, such as collagen, which behaves as

extracellular matrix (ECM) to build 3D structures for long-term culture

(similar to the 3D scaffold used in previous method), is generally squeezed out

[139] from the nozzle rather than ejected by a printhead. The reason why not

to directly print the thermo-reversible gel, or even straightforward, the cell-


69 of 176

laden gel is that the conventional piezo-based (or thermal-based) inkjet

printhead is not easily adapted for solutions with high viscosity such as

collagen. Solutions with quite low concentration of collagen (normally less

than 0.5%, w/v), certainly, can be successfully printed, but is undesired when

acting as extracellular matrix. To overcome this limitation, one method is to

design and fabricate inkjet printhead that capable of printing liquids with high

viscosities [140]. Another method is to utilize liquid materials that capable of

solidifying after ejection, such as sodium alginate (SA) solution and

fibrinogen solution [141]. For example, cell-laden alginate can be printed onto

substrate which contains calcium chloride solution. Crosslink happens once

the two liquids meet together, and the crosslinked alginate is able to control

the position of the ejected cells.

Besides printing of cells, inkjet is also used to print different solutions [101],

polymers [142], proteins [143-145] and growth factors into the living-

structures, for mediating cell viability.

2.4.2 Challenges and Requirements

Although it has been widely acknowledged that “organ printing” is a

promising technology for creating 3D living structures, plenty of challenges

are still there and successful organ printing should most importantly fulfill the

following requirements:

Cells in suspensions tend to agglomerate over around half an hour, which will

lead to the non-uniformity in the cell ink. Moreover, bulks of cells can easily


70 of 176

clog the nozzle of the inkjet printhead and interrupt the printing process. Thus

either the organ printing process should be rapid enough, or devices that can

eliminate cell agglomeration should be designed and equipped into the cell ink

chamber.

Despite the high resolution provided by inkjet printing method, it is difficult to

exactly control a number of cells in one droplet, especially when small nozzle

and/or low-concentration cell ink is used. To eliminate “empty droplet” and

cell-less droplets, cell concentration in the ink, or the ink concentration, should

sufficiently high.

Furthermore, cells should get rapid and continuous deposition and

solidification onto the thermo-reversible gel; the gel must provide adequate

mechanical support to the living structures; the gel should also allow adhesion,

spreading and proliferation of multiple cell types; sufficient oxygen and

nutrients have to be supplied to the cells which deep within the 3D structures;

the gel should degrade in a regular and predictable fashion; mechanical

strength of the grown-up tissue should also be considered as huge structure

might not be strong enough to hold together by itself once the gel is removed.

To conclude, despite the various existing challenges, “organ printing” is

feasible, fast-developing and predicted to be one of the most promising

technologies in tissue engineering. It uses the principle of cellular self-

assembly into tissues [146] and aims to build implantable organs to treat

diverse diseases such as cancer, loss of tissue function, or organ failure. “It is


71 of 176

safe to predict that in the 21st century, cell and organ printers will be as

broadly used as biomedical research tools as was the electron microscope in

the 20th

century”. [137]

Chapter 3: Novel Printhead Design

72 of 176

3. NOVEL PRINTHEAD DESIGN

The design and fabrication of a PET/PTFE-based piezoelectric squeeze-mode

DOD inkjet printhead with interchangeable nozzles is presented in this section.

The two methods, heating and pulling glass tubing, and silicon micro-

machining, which are used to fabricate nozzles for the printhead, will also be

reported. The characteristics of this novel printhead are studied by dispensing

glycerin-water solutions and non-Newtonian sodium alginate (SA) solutions,

and the experiment results with discussions are also documented in this

section.

3.1 Introduction

As has been mentioned before, due to its unique advantages, the application of

DOD inkjet printing technology has been expanded from conventional graphic

printing to new areas, such as fabrication of integrated circuits (ICs) [20, 147],

LED [44], rapid prototyping (RP) [148], MEMS, cell printing [2, 8, 9] and

drug delivery [149]. Accordingly, the dispensed liquids have been expanded

from the conventional pigmented ink (or standard dye-based ink) to polymers

[12, 150-152], gels, cell ink or other materials which often have higher

viscosities or even contain large particles or cells. Consequently, the

traditional inkjet printer designed for graphic printing is unable to fulfill the

new challenges, one of which is to dispense fluids of very high viscosities. For

most of the commercial inkjet printheads supplied by companies like

Microdrop, Microfab, Dimatix and XAAR, only liquids with viscosities lower

than 20 cps [12] can be consistently dispensed. Fluids with even higher


73 of 176

viscosities have to be diluted before printing or warmed up during the printing,

which will adversely affect the properties of the liquids.

Another challenge is raised by nozzle clogging. Fluids containing particles, or

cells, can easily block the nozzle orifice, resulting in time-consuming nozzle

cleaning or even damage of the entire conventional printhead. To solve the

problem, the easiest way is to use a nozzle with a bigger orifice, as bigger

orifices are less likely to clog. However, this is often not desirable in inkjet

printing as bigger nozzles result in bigger droplets and lower printing

resolution. In [21], Chen and Basaran reported that by judiciously controlling

the piezoelectric parameters governing the flow within the nozzle and thereby

the drop formation, droplets with diameters less than 40% of the orifice

diameter could be produced. A similar study was carried out by Goghari and

Chandra [22]. These studies reveal a possible way to solve this nozzle

clogging problem without sacrificing printing resolution. However, their

methods only work over a limited range of Ohnesorge numbers.

The poor printability and nozzle clogging may result in unreliable or failed

dispensing when using the traditional inkjet printhead design for complex

liquids.

In this section, we will present an in-house-developed PET/PTFE-based

piezoelectric squeeze mode inkjet printhead with an interchangeable nozzle

design. PET (polyethylene terephthalate) tubing, comprising of a much softer

material, is used as the printhead chamber to substitute for the conventionally


74 of 176

used glass tubing [140]. Liquids with viscosities of up to 100 cps have been

successfully dispensed by this novel printhead. When strongly corrosive inks

are involved, Teflon tubing is served as the printhead chamber. The

interchangeable nozzle design allows one to easily clean or change the

clogged or damaged nozzle, avoiding the destruction of the whole printhead

assembly.

3.2 Printhead Fabrication

Fig. 2.25 schematically shows the construction of a traditional piezoelectric

squeeze mode printhead. By using epoxy adhesive, a piezoelectric element is

tightly attached onto a glass tube which with an orifice at one end. When an

electrical pulse is applied, the piezoelectric element will contract inward,

squeezing the glass tube as well as the liquid inside. In order to eject a droplet

from the orifice, the volume change within the piezoelectric transducer, due to

the electrical pulse, must exceeds the volume of liquid to be ejected.

Furthermore, the volume change must be sufficient to develop enough

pressure inside the liquid to overcome the surface tension at the orifice. The

fractional volume change due to the piezoelectric effect is approximately:

(3.1)

where d31 is the piezoelectric strain constant, U is the applied voltage and t is

the thickness of the piezoelectric tube [61]. The negative sign indicates

contraction when the applied pulse has the same polarity as the original

polarizing voltage for the piezoelectric element. Equation 3.1 shows that the

printability of a printhead is mainly depended on the piezoelectric strain

constant and the geometry of the piezoelectric transducer. In this study, we


75 of 176

focus on how to improve the printability of piezoelectric squeeze printheads

without pursuing high piezoelectric strain constant. All the piezoceramic tubes

(PZT-5H, from Boston Piezo-Optics Inc.) have the same piezoelectric strain

constant d31, approximately -275×10-12

m/V, at 25 ˚C.

The basic idea is to reduce the energy loss during the deformation of the liquid

chamber, by replacing the traditionally used glass tube with PET or Teflon

tube. Accordingly, the printhead is divided into two parts: a printhead chamber

and an interchangeable nozzle attachment fitted tightly to the chamber by

screw threading. These will now be described in turn.

3.2.1 Printhead Chamber

The design of the printhead chamber is illustrated in Fig. 3.1. The PET heat-

shrink tubing has a relative low shrinking temperature ranges from 85 ˚C to

190 ˚C. Thus a hair drier is recommended to be the heat source, rather than a

burner which could burn up the tubing if overheated. To get a uniform

shrunken tubing with a desired diameter, a steel tube with 4.9 mm OD is

inserted into the PET tubing during the heating process, as a mould. The PET

tube with 6.0 mm OD and 0.1 mm wall thickness (230400CHGS, from

Advanced Polymers, Inc.) is evenly heated, shrinking it to a tubing with

approximately 5.2 mm OD, so that it can fit exactly inside the piezoceramic

tube. This shrunken PET tubing is used as the inner wall for the printhead

chamber, which directly contacts with the liquid to be dispensed. By using

electrical conductive epoxy (CW2400, from ITW Chemtronics Inc.), the

shrunken PET tubing is glued to the inner wall of the piezoceramic tube (PZT-


76 of 176

5H, from Boston Piezo-Optics Inc.) with 6.35 mm OD, 0.5 mm wall thickness

and 25.4 mm in length.

Fig. 3.1: The novel printhead. (a) Schematic showing of the design (out of

proportion). (b) A self-fabricated printhead following the novel design.

Teflon tubing serves as the printhead chamber when strongly corrosive inks

are involved, due to its perfect anti-corrsive property; however, it is such a

non-stick material to be directly bonded to the inner surface of the

piezoceramic tube. Fortunately, sodium-based chemical etchant can be used to

etch the surfaces of the Teflon material, to make it bondable to another

material. In this study, the PrimeEtch® Plus solution, provided by Plastomer

Technologies, an EnPro Industries company, is used as the etchant. Teflon

tubing (from Zeus, Inc.) with 5.22 mm OD and 0.25 mm wall thickness is

dipped into the etchant for 5 minutes. The etching takes place to a depth of a


77 of 176

few hundred angstroms and modifies only the surface composition of the

Teflon tubing, leaving other properties of the tubing unaffected, even the

dimensions. The etched Teflon tube is then rinsed in alcohol for 2 minutes,

dried, and glued to the inner wall of the piezoceramic tube by using electrical

conductive epoxy (CW2400, from ITW Chemtronics Inc.), forming the

printhead chamber.

Fig. 3.2: Schematic showing the fabrication of the printhead chamber: (a) PET tube

before shrink. (b) Teflon tube before etching. (c) The steel tube used as a mould

during heating of PET. (d) PET tube after shrink. (e) Teflon tube after etching. (f)

Piezoelectric tube. (g) Shrunken PET tube bonded to the piezoelectric tube.

In the next step, two wires are separately attached to the inner and outer wall

of the piezoceramic tube by using electrical conductive epoxy, for connecting

the printhead to the piezo–driver, as shown in Fig. 3.2(g). Then the whole part

is fixed inside a brass housing by araldite epoxy adhesive for protection. The

solidified araldite epoxy can also prevent short circuit which can be caused by

liquid permeation to the piezoceramic tube. The two connecting wires are

pulled out through a hole in the housing, and the hole is also sealed with


78 of 176

araldite epoxy, as shown in Fig. 3.1(b). An outside thread is cut on the bottom

of the brass housing for connecting the housing to the nozzle adaptor. Fig. 3.3

schematically shows the design of the brass housing and the nozzle adaptor.

Fig. 3.3: Schematic showing the design of the printhead housing and the nozzle

adaptor.

3.2.2 Interchangeable Nozzle Design

The glass nozzle is fabricated by heating and pulling a glass tube, as

demonstrated by Lee [68]. The setup is graphically shown in Fig. 3.4(a). A

glass tube with 5.0 mm OD and 3.5 mm ID is vertically fixed to a motor which

rotates the tube about its axis. By applying local heat to the lower section of

the rotating glass with a propane torch, the glass tube is melted at the location

of the flame, and pulled longer by the weight of its lower portion until it


79 of 176

finally breaks into two parts, each of which contains a hollow cone with a

closed end. The closed end is then polished by fine sand papers until an orifice

of a desired diameter is exposed, as shown in Fig. 3.4(c). By this method,

orifices of 13 µm to 300 µm have been fabricated in this study. A similar

glass-fabricated nozzle was also adopted by Fan et al. [153].

Fig. 3.4: Fabrication of a glass nozzle by heating and pulling glass tubing. (a)

Drawing of the glass tubing heating system (out of proportion). (b) Glass tubing

containing a hollow cone with a closed end. (c) A 50 µm orifice fabricated by

polishing the end of the tubing showing in (b).

As recommended by Lee [68], to generate an axisymmetric conical nozzle

profile, the rotation speed of the motor is maintained around 600 rpm.


80 of 176

Fig. 3.5: Fabricating glass nozzle by heating and pulling 1.0 mm glass capillary with

a micropipette puller. (a). The P-97 Flaming/Brown type micropipette puller. (b).

Heating the capillary. (c). Hit the sharp tip to from an orifice.

The major advantage of this nozzle fabrication method is ease of manufacture

and low cost. However, it is difficult to precisely control the taper angle of the

nozzle.


81 of 176

Fig. 3.6: Different shapes of tips fabricated by the micropipette puller. (a). A too

“sharp” tip. (b). A tip with a moderate converging shape.

To eliminate this limitation, a professional micropipette puller (the P-97

Flaming/Brown type micropipette puller, from Sutter Instrument Company)

can be used to heat and pull the glass capillary (TW100-4, from World

Precision Instruments) with 1.0 to 2.0 mm OD diameter. As shown in Fig.

3.5(b), a glass capillary is inserted through the heating element. Two screws

are used to fix the two ends of the capillary. Two separate springs are

connected to the screws, thus the screws will pull the two capillary ends to

opposite directions. After setting a desired program, the capillary is heated and

pulled, breaking into two parts, each of which contains a hollow cone with a

sharp tip. The sharp tip is then hit by a heated hitting wire under microscope,

forming a tiny orifice, as shown in Fig. 3.5(c).

Using this method, a sharp tip with an outer diameter of even 60 nm can be

fabricated. However, such a sharp tip normally has a slender shape, as shown

in Fig. 3.6(a). It is not suitable for inkjet printhead nozzle usage as it has too

much flow resistance, which will lead to difficult jetting even failure of jetting.

Actually, the shape of the sharp tip can be controlled by setting different

program, in other words, by changing the heating temperature and the pulling

speed. Fig. 3.6(b) shows a desired tip shape for inkjet printhead nozzle


82 of 176

application. The second tip has a moderate converging shape and will have a

lower flow resistance. However, in turn, leads to a relatively bigger orifice

diameter. A tip with an orifice diameter of 13 µm is shown in Fig. 3.7. Tip

with an orifice as small as 7 µm has been successfully adopted for jetting in

this study.

Fig. 3.7: A 13-micron-tip fabricated by the micropipette puller.

Fig. 3.8: Inkjet printhead nozzles fabricated from glass tube.

Fig. 3.1 also shows how the interchangeable nozzle design is implemented.

After being fixed to a short brass cylinder (Cap 2 in Fig. 3.1) by araldite epoxy

adhesive, the nozzle is placed inside another brass cap (Cap 1 in Fig. 3.1) that

has an inside thread which is tightly fitted to the outside thread of the


83 of 176

printhead chamber. An O-ring must be used here to prevent cracking of the

nozzle from overtightening of the threads. Fig. 3.8 shows different glass

nozzles fabricated in this study.

Although tips with uniform shapes can be fabricated by this micropipette

puller, nozzles fabricated from such tips are normally quite fragile.

Furthermore, when it goes to the tip hitting process to form the orifice, non-

uniformity emerges again. To obtain mass productability and exact

reproducible desired nozzle profile for large production runs, silicon

micromachining method can be used [68].

3.3 Experimental Testing of the New Printhead

3.3.1 Experimental Setup

Experimental tests were carried out to investigate the characteristics and

repeatability of the PET/PTFE-based printhead, as well as to compare the

ejection capacity of the PET/PTFE-based and the glass-based printheads.

The experimental setup is comprised of an air compressor, a pressure regulator,

a liquid reservoir, a piezoelectric actuated printhead, a piezo driver, a

stroboscope light and a CCD camera, as shown in Fig. 3.9.

The fluid to be dispensed is filled into a 60 ml stainless steel reservoir which is

mounted on a XYZ motion stage. The combination of the air compressor and

the pressure regulator (AD 3000D, from Iwashita Instruments Pte Ltd.)


84 of 176

provides a negative pressure in the reservoir to hold up and prevent the liquid

from leaking out of the orifice of the printhead. Before getting into the

printhead, the liquid is filtered by passing it through a syringe filter with 0.2

µm pore size membrane (Ref. 4652, from Pall Corporation) to remove large

particles which might block the nozzle. Electric signals are sent by a

JetDriveTM

III (from Microfab Technologies Inc.) to the piezoelectric

transducer, causing alternating expansion and contraction of the transducer as

well as the printhead chamber, ultimately, squeezing the liquid inside the

chamber and ejecting a droplet from the orifice.

The inkjet process is produced by a periodic driving voltage and the resulting

droplet ejection is repeatable from one droplet to the next. This allows for

stroboscopic imaging to determine the formation and the ejection velocity of

the droplets. Concurrently with the printing, signals are also sent by the driver

to a stroboscope (MS-200, from Nissin Electronic Co., Ltd.) which has pulse

duration of 2 s and is therefore capable of freezing an image of the high-

speed droplet with minimum blur. The droplet shape is illuminated by the

flashing of the strobe light and the images are captured by a JAI CV-A11

camera (from Ultravision Pte Ltd.). To determine the droplet velocity, the

stroboscope was operated at the same frequency as the printhead driver. By

varying the time delay between the signal for the stroboscope and signal for

the piezoelectric transducer, sequential images of the droplet during its motion

are captured with known time differences. The droplet velocity can then be

derived by dividing the spacing between the droplets by the time difference


85 of 176

between two frames. Fig. 3.10 shows a representative typical droplet

formation sequence, the droplet is 50 µm and its velocity is 0.69 m/s.

Fig. 3.9: Schematic showing of the drop-on-demand inkjet printing system used in

the experiment.

Fig. 3.10: Image sequences showing the formation of a 50 µm droplet from a 36 µm

inkjet nozzle. The times shown are 0, 144, 322, 367, 389, 400, 522 and 1122 µs

relative to the first frame. The droplet velocity is here determined to be 0.69 m/s.

The droplet size is determined by two different methods, i.e. either measured

directly from the images, or by a weight method. In the latter method, printing

is carried out in a vapor-saturated environment to suppress the evaporation of


86 of 176

the printed liquid, and then 7200 droplets are collected and then weighed for

each test condition. Droplet diameter can then be calculated from its weight

and the corresponding liquid density.

3.3.2 Experimental Conditions

The jet driver made by Microfab Technologies Inc. is able to send out voltage

pulses with designed profiles. Up to 12 points can be set to form the signal

waveform. Commonly used signals are of uni-polar, bi-polar or sinusoidal

shape. The maximum allowable amplitude and frequency for the pulse is ±140

V and 30 kHz, respectively.

Fig. 3.11: Schematic showing of the uni-polar pulse waveform.

Fig. 3.11 shows a uni-polar pulse employed in the experimental study. The

zero line represents the equilibrium state of the piezoceramic tubing, without

any external voltage. During the time of trise, the piezoceramic tubing expands

outward to its maximum inner volume and holds that state for a time of tdwell.

During the time of tfall, the piezoceramic tubing contracts inward, to its

equilibrium state. The expansion and contraction of the piezoceramic tubing

causes negative and positive pressure waves propagating and reflecting inside

the printhead, which ultimately leads to droplet ejection [52]. During all the


87 of 176

experiments, trise and tfall were kept at 3 µs, which are the minimum permitted

values of the jet driver, with the purpose of introducing instant action of the

piezoeceramic tubing.

Static pressure needs to be applied to the reservoir, so that the liquid will not

flow out of the nozzle under the hydrostatic pressure. The negative pressure

applied to the reservoir was determined by direct observation showing no

liquid leaking and no air entertainment.

3.3.3 Testing Liquids

Fig. 3.12: Measured viscosities for different concentrations of sodium alginate

solutions. Measurement at 20 ˚C.

The conventional pigmented ink or standard dye-based ink for graphic printing

normally has a viscosity of less than 5 cps. However, to apply inkjet printing

in the new areas mentioned earlier, various complex liquids like polymers,

gels and other materials with much higher viscosities need to be effectively


88 of 176

dispensed. Thus, two types of liquids are used to test the new printhead:

aqueous glycerin solutions with viscosities from 1 to 120 cps, and aqueous

sodium alginate (A2158, low viscosity, Sigma-Aldrich) solutions with

concentrations from 0.2% to 2.8% (w/v).

Aqueous solutions of sodium alginate were prepared by suspending the

polymer in distilled water. After 6 hours of stirring by a magnetic stirrer, the

solution was sterilized by sterile filtration, using 0.2 µm pore size membrane

filters. The filtration did not change the concentration of the solution as the

polymer has been totally dissolved.

The viscosity of the sodium alginate solutions was measured using the ARES

Rheometer (TA Instruments, Inc.). The geometry involved consists of two

parallel plates of 50 mm diameter with a gap of 0.5 mm. 1 ml sample was used

for each measurement. The shear rates ranged from 1.0 to 4000 s-1

. As shown

in Fig. 3.12, aqueous sodium alginate (A2158, low viscosity, Sigma-Aldrich)

solutions behave as Newtonian fluids at low concentration of 0.2% to 1.6%

(w/v) as the corresponding viscosities remain almost constant over a wide

range of shear rates (1.0 to 4000 s-1

). The solution shows significant shear

thinning behavior once the concentration exceeds 1.8% (w/v).


89 of 176

3.4 Experimental Results

3.4.1 Comparison of PET/PTFE-Based and Glass-

Based Printhead

Fig. 3.13: Threshold voltages for PET-based printhead (–○–), PTFE-based printhead

(–*–) and glass-based printhead (–■–). Nozzle diameter is 119 µm.

The commercial printheads fabricated by Dimatix, XAAR, Microfab and

Microdrop can only dispense liquids with viscosity less than 20 cps [12]. In

our printhead design, PET or Teflon tubing was used as the printhead chamber,

but a separate glass-based printhead of similar configuration was also

fabricated for this study to compare the dispensing capacity of the new type

printhead and that of the glass-based printhead. Aqueous glycerin solutions

with viscosities from 1 to 120 cps were used for the test. Fig. 3.13 shows the

threshold voltage needed for dispensing glycerin solutions of different


90 of 176

viscosities. During the printing, the same nozzle of 119 µm diameter was used

for both of the PET/PTFE-based and glass-based printheads.

It can be seen that the threshold voltage increases with the increase of fluid

viscosity, for all of the printheads. However, for the same viscosity, the

PET/PTFE-based printhead requires a much smaller threshold voltage than the

glass-based printhead. Furthermore, the PET/PTFE-based printhead can

dispense liquids with viscosity of up to 100 cps, which far exceeds the

performance of the glass-based ones which are typically used in commercial

printheads. The main reason for the lower voltage is that PET or Teflon is

much softer than glass. When an electrical pulse is applied, the liquid chamber

made of PET or Teflon tube is much easier to be squeezed by the

piezoceramic element, thus less energy will be dissipated in the deformation of

the liquid chamber. Consequently, a larger volumetric change will be achieved

in the liquid, leading to a better dispensing capacity.

Furthermore, from eq. 3.1, for the piezoelectric tubes of the same wall

thickness, the one with the larger diameter can generate a greater change in its

volume. This is the reason why the self-fabricating glass-based printhead can

only dispense glycerin solution with viscosity of 30 cps, which still exceeds

the limitation of most commercial printheads. However, when the diameter of

the piezoelectric tube increases, the diameter of the inner glass tube or

PET/PTFE tube should also increase. Glass material is quite stiff and brittle;

therefore, with the increase of its diameter, it becomes more fragile when the

same wall thickness is used. However, with a thicker wall, more energy is

absorbed in the glass and less volume change can be obtained. Fortunately,


91 of 176

PET or Teflon is quite pliable; by adopting them as the printhead chamber, the

piezoelectric tube with a bigger diameter can be used to further improve the

printhead behavior for dispensing materials with high viscosity.

3.4.2 Effect of Pulse Width

For PET-based printhead, the effects of pulse width on droplet diameter and

droplet velocity were investigated by keeping the pulse amplitude constant at

50 V. Note that here the pulse width represents the duration of tdwell in Fig.

3.11. The jetting frequency was kept constant at 120 Hz. The PET-based

printhead and PTFE-based printhead behave almost the same, for

simplification, only the results for PET-based printhead will be represented

and discussed.

Fig. 3.14: Effects of pulse width on droplet velocity and droplet size. The pulse

amplitude is 50 V. Nozzle diameter is 119 µm.


92 of 176

It is observed from Fig. 3.14 that both droplet velocity and droplet diameter

initially increase with the increase of pulse width. The maximum droplet

velocity 4.3 m/s is obtained when the pulse width reaches a value of 100 µs.

Then droplet velocity rapidly decreases with a further increase of pulse width,

reaching a minimum value of 1.2 m/s at 360 µs pulse width. The increase of

droplet diameter continues until the pulse width is 120 µs. Then the droplet

diameter is almost constant at a value of 188 µm until the pulse width reaches

270 µs, followed by a dramatic fall to the value of 135 µm. Both the droplet

velocity and droplet diameter reach a near constant value after the pulse width

exceeds 420 µs. With further increase of pulse width to 990 µs, air gets easily

sucked into the nozzle, leading to a slight decrease of droplet velocity and

droplet diameter.

According to De Jong et al. [154], there are two types or two reasons for this

air entrapment. In the first type, the air entrapment is often related to the

presence of an ink layer on the nozzle plate, especially for a nozzle plate

without hydrophobic treatment. Within this ink layer, dust particles, which are

deposited from the ambient air, can be transported toward the inkjet nozzle

[155]. The reason for this transport lies in the fact that: during the time trise and

tdwell, the piezoelectric element has expanded outward, resulting in a negative

pressure in the printhead which drives the ink (with the dust particles) toward

the nozzle. Furthermore, according to Beulen et al. [156], the jet of droplets

will transfer momentum to the ambient air, which will result in a suction of air

towards the jet. Friction between the ink layer and the air flow induced by the


93 of 176

jet will also cause a flow of ink towards the jet nozzle. This so-called pulled

back flow will also cause the transport of particles toward the inkjet nozzle.

When the particles reach the jetting nozzle, they will cause a local surface

tension distortion and thus an asymmetry of the droplet formation. This

asymmetry of the retracted meniscus in combination with the next symmetric

pressure wave then will cause air entrapment [154, 157]. The entrained tiny

bubbles will oscillate and coalesce with the neighbouring bubbles, moving

along the inner wall of the printhead and growing through coalescence and

rectified diffusion, forming a bigger bubble [158]. When this bubble grows big

enough, normally much bigger than the diameter of the nozzle, it will absorb

too much of the pressure energy generated by the piezoelectric element and

thus stop the jetting. From this point of view, at a higher jetting frequency,

more particles are likely to be driven toward the nozzle, thus air entrapment is

more likely to happen.

In the second type, the air entrapment is also related to the presence of an ink

layer on the nozzle plate. A void is formed once the meniscus is pulled back,

due to the expansion of the piezo element. At the meantime, the pulled back

ink from the ink layer closes the void, forming an air bubble inside the nozzle.

In this experiment, the nozzle plate does not have any hydrophobic treatment,

thus a thin layer of ink will exist on the nozzle plate during printing.

Furthermore, the jetting frequency was kept constant at 120 Hz here, while the

air entrapment only happened when the pulse width exceeded 990 µs. Thus we


94 of 176

would like to say that the air suction in this experiment belongs to the second

type of air entrapment. When a waveform of a very big pulse width is applied,

the piezoelectric element expands for a longer time. As a result, meniscus is

pulled too far into the nozzle during each dispensing process, allowing air to

be easily sucked into the nozzle, forming a void inside the nozzle. At the same

time, the pulled back ink from the ink layer on the nozzle plate closes the void,

forming an air bubble inside the nozzle. Finally, a much bigger air bubble is

formed inside the nozzle and totally stops the dispensing at a pulse width of

1160 µs. Single droplets without satellites can now be obtained in two ranges

of pulse widths: 25 µs to 32 µs, and 420 µs to 1130 µs.

3.4.3 Effects of Voltage Pulse Amplitude

Fig. 3.15: Effects of pulse amplitude on droplet velocity and droplet size. The pulse

width is 100 µs. Nozzle diameter is 119 µm.


95 of 176

The effects of the pulse amplitude on the behavior of the PET-based printhead

were investigated by dispensing 5 cps aqueous glycerin solution through a 119

µm nozzle. From Fig. 3.14, it can be seen that the optimal tdwell for generating

a high velocity droplet is 100 µs. Herein the tdwell was kept constant at 100 µs.

The pulse amplitude was varied to the maximum which can be generated by

the jet driver, i.e. 140 V. The jetting frequency was kept constant at 120 Hz.

As shown in Fig. 3.15, both drop velocity and drop volume increase initially

with an increase of pulse amplitude. However, single droplets can only be

obtained using pulse amplitudes from 21 to 40 V. Further increase in pulse

amplitude generates a primary droplet followed by a small satellite droplet.

This is understandable as a higher voltage causes a bigger volume change

within the piezoelectric element, thus a longer column of liquid squeezed out

and a satellite will be generated. The satellite droplet becomes bigger and

tends to break into multiple satellite droplets as the pulse amplitude is further

increased. The maximum droplet velocity is 3.24 m/s and the droplet diameter

varies from 150 µm to 200 µm.

A slight decrease in droplet velocity is observed once pulse amplitude exceeds

76 V. The reason for this decrease lies in the fact that: during the time trise and

tdwell, the piezoelectric element has expanded outward, resulting in a negative

pressure in the printhead which causes the meniscus to move into the nozzle.

When very high voltage is applied, the piezoelectric element expands more.

As a result, meniscus is pulled too far into the nozzle during each dispensing

process, allowing air to be easily sucked into the nozzle. The sucked air forms


96 of 176

small air bubbles inside the nozzle which leads to the decrease of droplet

velocity [159].

3.4.4 Nozzle Size

For a specific printhead with a fixed nozzle size, it is of interest to determine

the smallest droplet and the biggest droplet that can be generated.

Eight different nozzle sizes were investigated for this printhead. For each

nozzle size, the pulse width was 22 µs. Then the pulse amplitude was slowly

increased from 10 V to 140 V in steps of 1.0 V, until the liquid can be

regularly dispensed. The corresponding droplet diameter is recorded as

indicating for the smallest single droplet. The pulse amplitude was further

increased, until satellite droplet was generated along with the main droplet.

Then the pulse amplitude was decreased by 1 V step length and droplets were

collected. Droplet diameter was calculated and recorded as that for the biggest

regular droplets, i.e. without a satellite.

To determine the largest droplet, irrespective whether a satellite was produced,

a different approach was used. In accordance with the results of Fig. 3.14 and

Fig. 3.15, to obtain the biggest droplet diameter, the pulse amplitude was set to

be 140 V. The pulse width was increased from 25 µs to 625 µs at intervals of

50 µs. Droplets were collected and weighed to determine their size. The

biggest value among the 13 samples was recorded as the biggest droplet

diameter.


97 of 176

Fig. 3.16 shows the relationship between the droplet size and the nozzle size,

for all nozzles tested. It is shown that all the above three types of droplet sizes

increase monotonically with increasing nozzle sizes. The droplets are always

larger than the nozzle diameter indicated by the broken line. An excellent rule-

of-thumb states that the droplets are between 120 to 220% of the nozzle

diameter.

Fig. 3.16: Effects of nozzle size on droplet diameter. (–*–) denotes the diameters of

the smallest single droplets can be generated; (–■–) denotes the diameters of the

biggest single droplets can be generated; (–▲–) denotes the diameters of the biggest

droplets which can be generated using the maximum voltage.

3.4.5 Repeatability

A good inkjet printhead should have nice repeatability, allowing generation of

a rapid sequence of droplets without big variation in droplet velocity and

droplet size. To test the repeatability of our novel PET-based printhead, water

was dispensed through a 119 µm nozzle. In accordance with Fig. 3.14, to

generate stable single droplets, the printing was carried out by using a signal


98 of 176

of 50 V pulse amplitude and 600 µs pulse width. As shown in Fig. 3.17, over a

one hour test run at 120 Hz, the variation in droplet diameter is only 0.18%

and 0.46% for droplet velocity.

Fig. 3.17: Repeatability test of the PET-based printhead. Nozzle diameter is 119 µm.

3.4.6 Maximum Jetting Frequency

As a manufacturing tool, high speed jetting is required to increase productivity

of inkjet printing technology. For industrial printer with multi-nozzle, this can

be realized by increasing number of nozzles or increasing jetting frequency of

each nozzle. While for printhead with single nozzle, as designed in this study

(also for Microfab and Microdrop), jetting speed can only be improved by

increasing jetting frequency. However, for a reliable jetting, a subsequent

droplet should not be ejected until the pressure wave from the previous pulse

signal has sufficiently damped. This damping takes time and thus limits the

maximum jetting frequency [38]. For a specific printhead, its maximum jetting

frequency is mainly depended on the construction of the printhead as well as


99 of 176

the driving signal [40]. Typical DOD printheads generate droplets at rates in

the range 0.1-10 kHz.

Fig. 3.18: Effects of jetting frequency on droplet velocity and droplet size. The pulse

width is 100 µs. The pulse amplitude is 30 V. Nozzle diameter is 119 µm.

If a droplet is ejected before the pressure waves from the previous pulse signal

have sufficiently damped, the new droplet ejection cycle will be affected by

the non-zero flow field inside the printhead. Consequently, the droplet velocity

and the droplet size will increase or decrease, depending on whether the

residual movement of the meniscus is in-phase or out-of-phase with the new

droplet ejection cycle [160].

Fig. 3.18 shows the effects of jetting frequency on droplet velocity and droplet

diameter, for the PET-based printhead. Printing was carried out by dispensing

water through a 119 µm nozzle. The tdwell was kept constant at 100 µs, the

optimal value for the printhead. The pulse amplitude was kept constant at 30 V.


100 of 176

The printing was operated with a variation in frequency between 1 Hz to 5

kHz. It is shown that below 1.5 kHz, jetting frequency has relatively small

effects on drop velocity and drop diameter. The reason was that below this

frequency, there was sufficient time (670 µs, according to a jetting frequency

of 1.5 kHz) between droplet ejection cycles for the acoustic pressure waves to

get damped. Thus the droplet ejection cycles were independent of each other

and were irrelevant with the jetting frequency.

However, when jetting frequency exceeds 1.5 kHz, both drop velocity and

drop volume rapidly increase with an increase of jetting frequency. The

maximum droplet velocity 5.7 m/s is obtained when the jetting frequency

reaches a value of 2.1 kHz. Then droplet velocity decreases with a further

increase of jetting frequency. The increase of droplet diameter continues until

the jetting frequency is 2.3 kHz. Then the droplet diameter also decreases with

a further increase of jetting frequency. The dispensing is stopped at a

frequency of 3.7 kHz. The maximum jetting frequency of 3.6 kHz is higher

than that of the Microdrop printhead (2.0 kHz), while much lower than that of

the Dimatix, XAAR, and Microfab printhead (20 kHz).

The strong variation of droplet velocity and droplet diameter with changing of

jetting frequency indicates that, above 1.5 kHz, the time interval between two

consecutive droplet ejection cycles was not sufficiently long for the acoustic

pressure waves to get adequately damped. With the meniscus motion method

proposed by Kwon [161], one can estimate the time needed for the damping of

the pressure waves. For the printhead designed in our study, it has an optimal


101 of 176

tdwell around 100 µs; and it takes around 800 to 1000 µs for the acoustic

pressure waves to get sufficiently damped. As a result, our printhead should

has a much lower threshold frequency [160], above which jetting frequency

will have great effects on droplet speed as well as droplet volume. This is

verified by the experiment results, as shown in Fig. 3.18. The threshold

frequency for the designed printhead is only 1.5 kHz.

The maximum droplet velocity is produced with a driving frequency of around

2.1 kHz, corresponding to the resonance frequency of the inkjet channel. This

resonance frequency is much lower as compared to other commercial

printheads. The reason is that the first mode resonance frequency of the

printhead is inversely proportional to the length of the pressure channel [160];

in the meantime, the designed printhead has a liquid channel of 50 mm length,

which is much larger than that of the commercial printheads. As a result, a

much lower resonance frequency is reasonably expected.

3.4.7 Jetting of Non-Newtonian Liquid

Aqueous sodium alginate (SA) solutions were used to estimate the printing

behavior for non-Newtonian liquid of our PET-based printhead. The threshold

voltages for different concentrations of SA solutions are compared with those

values for glycerin solutions (which have been shown in Fig. 3.13). To do the

comparison, the viscosity used for the non-Newtonian SA solutions is that

predicted from Fig. 3.12, corresponding to the mean shear rate of 1.0×104 s

-1.

The selection of this mean shear rate to represent the real values is reasonable.

The reason is that the nozzle used in the test has a diameter of 119 µm, and the


102 of 176

velocity of the liquid thread ejected from the nozzle is around 1.5 to 3.0 m/s

before the droplet formation. Thus the mean shear rate determined by ,

where v and r are the drop velocity and radius, respectively, does have the

order of 104

s-1

.

Fig. 3.19: Threshold voltages for sodium alginate solutions of concentrations from

0.2% to 2.8% (w/v).

Fig. 3.19 shows that for SA solutions, the threshold voltage increases from 13

V to 110 V when the SA concentration increases from 0.2% to 2.8% (w/v). In

the initial stage, for the same viscosity, the threshold voltage required by the

SA solution and glycerin solution is nearly the same. However, when the

concentration of the SA solution further increases to 2.0% (w/v), the SA

solution needs a larger threshold voltage than the glycerin solution having the

same viscosity, especially when the SA solution concentration exceeds 2.4%

(w/v). It is noticeable that although 110 cps glycerin solution is dispensable,

the dispensing of 3.0 % (w/v) SA solution with a viscosity of 63 cps failed.


103 of 176

There is a big jump of threshold pulse amplitude from 58 V for 2.6% (w/v) SA

solution to 134 V for 2.8% (w/v) SA solution, despite that the viscosity

difference is only around 4 cps.

Fig. 3.20: Schematic showing of drop formation for 2.2% SA solutions.

The results indicate that during the inkjet printing, sodium alginate solutions

behave more like a Newtonian fluid when its concentration is below 2.0%

(w/v). However, with the increase of concentration, sodium alginate exhibits

more non-Newtonian behavior. The dispensing of non-Newtonian liquid

generates longer threads than the Newtonian fluids with similar viscosity [65],

which has been observed from the experiment. A droplet formation for 2.2%

(w/v) SA solution is shown in Fig. 3.20. It is shown that a quite long liquid

thread is formed before the droplet separation from the nozzle. The breakup of

the non-Newtonian liquid thread causes much energy dissipation due to the

elasticity in the fluid, leading to a significantly lower droplet velocity. For the

3.0% (w/v) SA solution, energy dissipation is so much that the dispensed


104 of 176

liquid does not have enough kinetic energy to overcome the surface energy,

leading to the failure of droplet separation.

It also can be seen that the long liquid thread snaps back and combines with

the primary droplet, rather than pinching off from the primary droplet and

breaking up into a satellite droplet. The phenomenon verifies the previous

claim that an increase in elasticity of the solution will effectively eliminate

satellite generation [65].

3.5 Conclusions

A PET/PTFE-based piezoelectric DOD inkjet printhead with an

interchangeable nozzle design has been proposed and fabricated by the authors.

The printhead chamber is made of PET or Teflon tube, which is much softer

than the commonly used glass tube. The ejecting capacity of this novel

printhead has been compared with commercial printheads, and found to have

superior performance and versatility. Aqueous glycerin solutions with

viscosity as high as 100 cps have been successfully dispensed, while the

corresponding commercial printheads can only dispense liquids with

viscosities lower than 20 cps. PTFE-based printhead provides excellent anti-

corrosive property when strongly corrosive inks are involved. The

interchangeable nozzle design largely alleviates the difficulty in cleaning of

clogged nozzles and greatly reduces the occurrence of printhead damage. One

of the printhead which was fabricated in 2007 still works properly now The

effects of operating parameters, including voltage pulse amplitude, pulse


105 of 176

width and jetting frequency, on droplet size and droplet velocity have been

characterized. The new printhead shows excellent repeatability.

Finally, non-Newtonian aqueous sodium alginate (SA) solutions with

concentrations from 0.2% to 2.8% (w/v) have been successfully dispensed.

For relatively low concentrations, the threshold voltages required by SA

solutions and glycerin solutions take nearly the same values, implying that the

printhead characteristics calibrated from dispensing Newtonian liquid can also

be used as a reference to predict dispensing of the non-Newtonian liquids.

However, for higher concentrations of the polymers there is a sharp transition

where printing can no longer be achieved.

Chapter 4: Forming A Fine Jet In Inkjet Printing

106 of 176

4. FORMING A FINE JET IN INKJET

PRINTING

The formation of fine jet during the piezoelectric DOD inkjet printing will be

presented in this section. An ultra-high-speed video camera is used to record

the fast jetting process. The characteristics of this fine jet are studied by

dispensing different concentrations of aqueous glycerin solutions, and the

experimental results are presented and discussed in this section.

4.1 Introduction

Jet eruptions from free liquid surface can be found in a number of flow

configurations. These include Worthington jets [162, 163], which are

generated by the collapse of an impact crater; granular jets, discovered by

Thoroddsen and Shen [164], which are generated by the impact of a solid

sphere onto a deep bulk of granular material [165, 166]; the Cavity jets [167,

168], which are produced inside cavitation bubbles when they collapse and are

capable of severely damaging a solid surface, such as the blades of turbines

and kidney stones; apex jets [169], and many other kinds of jets [170, 171, 172,

173].

Over the years, there has been a consensus that jet formation in many different

cases is related to a singularity on a free surface [174, 175]. Fig. 4.1 [168]

shows a micro-jet formed by impact of a tube filled with perfectly wetting

liquid, on a rigid floor. In the gravity-free reference frame, when the tube falls


107 of 176

axially under its own weight, surface tension deforms its liquid interface into a

hemispherical shape. It is believed that the jet is generated by the violent

reversion of the interface curvature resulting from the impact. This is a typical

example to show that the jet formation is closely related to a singularity on a

free surface. Now it is quite interesting to us whether such needle-like jet can

also be produced by DOD inkjet printing, as the similar processes which

existing in a typical droplet ejection cycle: firstly an inwards hemispherical

meniscus is produced by the expansion of the piezoceramic element, then a

positive acoustic pressure generated by the contraction of the piezoceramic

element suddenly squeezes the liquid inside the printhead chamber, behaving

as the impact in Fig. 4.1. This is the part which will to be covered in the next

two subsections.

Fig. 4.1: Jet formation observed just after impact of the tube with a solid wall when

the free surface is initially deformed with a meniscus [168].


108 of 176

4.2 Experimental Setup

Printing experiments were carried out by using the self-developed squeeze

mode piezoelectric inkjet printing system. The setup is almost the same as that

shown in Fig. 3.9. The only difference is that an ultrahigh-speed video camera

(maximum frame rate: 1,000,000 fps) developed by Etoh et al. [176] is used to

record the jetting process here. Eight different concentrations of aqueous

glycerin solutions will be used as the ink, to obtain a wide range of viscosities.

4.3 Experimental Results

The jets are classified into two types: Type I is the typical fine jet, and Type II

is the jet produced when there is a bubble locating inside the nozzle.

4.3.1 Jet I

Fig. 4.2 shows a typical process of Type I jet. In the first 4 frames, a backflow

is generated by the expansion of the piezo-element and air gets sucked into the

nozzle due to the negative pressure produced inside the liquid. Surface tension

forces deform this liquid interface into a spherical shape, or a partial cavity.

The later sudden contraction of the piezo-element squeezes the fluid and

induces pressure gradients which in turn produce a sudden change in the liquid

velocity, and, finally causes the collapse of the cavity and the generation of the

fine jet. The principle is similar to that of the cavity jet shown in Fig. 4.1. Here

the differences between these two jets are the way of generating the free

surface (with singularity) and realizing the sudden change in the velocity field,


109 of 176

as well as the geometry of the liquid channels. Fig. 4.3 shows the converged

inner profile of the nozzle used for the experiments to study the fine jet.

Fig. 4.2: A 93 µm jet with a velocity of 7 m/s. The diameter of the orifice is 150

µm. Liquid used is 70% aqueous glycerin (w/w) solution. Printing parameters: bi-

polar piezo-driving signal with tdwell and techo equal to 700 µs; driving pulse amplitude

equals to 140 V. Negative pressure inside the reservoir is -2.2 kPa relative to the


temperature is 25 ˚C.

As compared to the regular printing cycle shown in Fig. 3.10, here the time

durations for the expansion and contraction of the piezoelement are much

longer of around 700 µs. The jetting is found to be extremely sensitive to the

negative pressure in the reservoir which is used to hold up and prevent the

liquid from leaking out of the orifice of the printhead. Fig. 4.4 shows another

jet process almost under the same conditions as that one shown in Fig. 4.2. As

can be clearly seen, by slightly changing the negative pressure, a much finer

jet with a much higher velocity can be produced.


110 of 176

Fig. 4.3: The 150 µm nozzle used for fine jetting experiments. The scale bar is 2 mm.

This image was taken when the nozzle was placed inside a 60% aqueous glycerin

(w/w) solution, which had an index of refraction similar to that of the glass.

Fig. 4.4: An 8 µm jet with a velocity of 29 m/s. is 150 µm. The liquid used is

70% aqueous glycerin (w/w) solution. Printing parameters: bi-polar piezo-driving

signal with tdwell and techo equal to 700 µs; driving pulse amplitude equals to 140 V.

The negative pressure inside the reservoir is -2.3 kPa relative to the atmospheric

pressure. Images were taken at a frame rate of 165 kfps. Ambient temperature is 25

˚C. The scale bar is 500 µm.

Fig. 4.4 also clearly shows the evolution of the cavity during the jet formation

process, as a much higher frame rate is used. Frames 1 and 2 record the

growth to maximum size of the cavity, as the liquid goes into tension due to


111 of 176

the expansion of the piezo-element. When the positive acoustic pressure

arrives, the cavity rapidly collapses towards the free surface (start from frame

3 to frame 16), followed immediately by the formation of a downward thin

liquid jet which rapidly grows in length (frame 17). It can be seen that the

shock flattens the lower surface of the cavity. Similar jet behaviors have been

also reported by Barrow et al. [177], in their study of cavitation damage.

4.3.2 Type II Jetting from Entrained Bubble

Fig. 4.5: A 16 µm jet with a velocity of 35 m/s. is 150 µm. The liquid used is


signal with tdwell and techo equal to 700 µs; driving pulse amplitude equals to 140 V.

Negative pressure inside the reservoir is -2.3 kPa relative to the atmospheric pressure.


The jet in Fig. 4.2 and Fig. 4.4 result from the collapse of a hemispheric crater.

In some cases, there is already an entrained bubble inside the nozzle before the


112 of 176

actuation of the piezo-element, the second type of jet will occur. This earlier

bubble is entrained in the late stage of the previous droplet ejection cycle, due

to the long duration of the piezo expansion. As shown in Fig. 4.5, the

preexisting bubble coalesces with the new piezo-generated hemispheric crater,

forming a new cavity which finally collapses and produces the fine jet.



signal with 450 µs tdwell and 70 µs techo; driving pulse amplitude equals to 140 V. The

negative pressure inside the reservoir is -2.3 kPa relative to the atmospheric pressure.

Images were taken at a frame rate of 27 kfps. Ambient temperature is 25 ˚C. The

scale bar is 500 µm.

Fig. 4.6 provides a close view for the cavities evolution inside the nozzle.

Frames 1 to frame 11 (a time duration of around 370 µs) record the growth to

maximum size of the piezo-generated cavity, as the liquid goes into tension

due to the expansion of the piezo-element. Concurrently, the preexisting


113 of 176

bubble is slightly pulled up. Starting from frame 10, the piezo-generated

cavity contacts and coalesces with the preexisting bubble. The neck

connecting the two cavities grows rapidly as studied by Thoroddsen et al.

[178]. The neck shapes form perfect circular arcs (marked by a blue arrow), as

demonstrated in Frame 12. A crest is developed when these circular arcs meet

the undisturbed bubble, as marked by the green arrow. With the growing of

the neck, this crest moves up the bubble and generates a series of capillary

waves, which propagate along the bubble surface and converge at the apex,

and finally, leading to the pinch-off of the small bubble (as marked by the

green circles in frame 13 and 14). A similar pinch-off phenomenon during

bubble coalescence has been comprehensively studied by Zhang and

Thoroddsen [179], who also provided a nice view of the capillary waves

focusing at the bubble apex, at a slight downwards angle. Finally, the shock

waves produced by the piezo-element arrive and cause the collapse of the

coalesced cavity, bringing about the fine jet. The jet is expelled out of the

nozzle by the positive acoustic pressures.


114 of 176


water. Printing parameters: bi-polar piezo-driving signal with 700 µs tdwell and 700 µs

techo; driving pulse amplitude equals to 140 V. The negative pressure inside the

reservoir is -2.3 kPa relative to the atmospheric pressure. Images were taken at a

frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 500 µm. (a).

The time interval between successive frames, dt, equals to 9.09 µs. (b). dt equals to

3.03 µs. (c). dt equals to 9.09 µs.


115 of 176



techo; driving pulse amplitude equals to 140 V. The negative pressure inside the


frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 500 µm. (a).

The time interval between successive frames, dt, equals to 9.09 µs. (b). dt equals to

3.03 µs. (c). dt equals to 9.09 µs.

Fig. 4.7 shows the details of the coalescence process with a very high frame

rate of 330 kfps. It is remarkable that we discovered that with the pinch-off of

the small bubble, a downward fine jet was produced simultaneously, as

marked by the arrows in Fig. 4.7(b). We conclude that this fine jet is different

from the one which produced by the cavity collapse. As shown in Fig. 4.2 and

Fig. 4.4, the cavity collapses when almost all the gas has been expulsed out of

the nozzle (frame 5 in Fig. 4.2); then the jet is produced (frame 6 in Fig. 4.2).

While here in Fig. 4.7(b), when this jet was produced (at frame 7 or frame 8)


116 of 176

inside nozzle, the big coalescence-generated cavity was still far away from

collapsing. Thus we can conclude that rather than the cavity collapse, both the

small bubble pinch-off and the jet formation are consequences of the local

surface collapse during the radial focusing flow, perhaps in a singular (self-

similar) fashion, as schematically shown in Fig. 4.8. Similar free surface

geometries evolution has also been reported for others’ study of cavity jet [180]

and impact jet [171, 173]. Here we can name this jet as the “surfaces collapse”

jet, to differentiate it from the cavity jet.

Cavity collapse occurs only after frame 6 in Fig. 4.7(c), followed immediately

by the cavity jet. Frame 8 in Fig. 4.7(c) shows that this cavity jet finally

catches up and collides with the first jet, forming a much thicker jet with a

complicated structure. Fig. 4.9 provides a much more clear view of this

collision.

Fig. 4.8: Schematic showing the free surface shapes.


117 of 176



techo; driving pulse amplitude equals to 140 V. Negative pressure inside the reservoir

is -2.3 kPa relative to the atmospheric pressure. Images were taken at a frame rate of

330 kfps. The numbers of the frames shown in the figure are n = 1, 4, 7 …… 52.

Ambient temperature is 25 ˚C. The scale bar is 500 µm.

In Fig. 4.6 to Fig. 4.9, it was shown that a small bubble can be pinched off

during the coalescence of the two cavities, for the 10% aqueous glycerin (w/w)

solution and pure water. Fig. 4.10 shows a very complicated jetting process for

the 70% aqueous glycerin (w/w) solution. From Fig. 4.10(b), we can conclude

that the two cavities did not coalesce when they got into contact with each

other, as a balloon-like bubble was generated here. So the process evolved as

following: a piezo-generated cavity was sucked into the nozzle and got into

contact with a preexisting bubble, without coalescing with it. The later piezo-

generated radial focusing flow pushed out the lower cavity as well as the

lower cap of the preexisting bubble, forming the balloon-like bubble extending


118 of 176

far out of the nozzle. This balloon-like bubble ruptures later close to its tip,

followed by a cavity jet which has been shown in Fig. 4.4.

Fig. 4.10: Images showing jetting produced when no coalescence happens between

the two cavities. is 150 µm. The liquid used is 70% aqueous glycerin (w/w)

solution. Printing parameters: bi-polar piezo-driving signal with 550 µs tdwell and 550

µs techo; driving pulse amplitude equals to 140 V. The negative pressure inside the


frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 500 µm. (a). dt

equals to 6.06 µs. (b). dt equals to 3.03 µs. (c). dt equals to 9.09 µs. (c). dt equals to

6.06 µs.


119 of 176

Fig. 4.10: Images showing jetting produced when no coalescence happens between

the two cavities. is 150 µm. The liquid used is 70% aqueous glycerin (w/w)




frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 500 µm. (a). dt

equals to 6.06 µs. (b). dt equals to 3.03 µs. (c). dt equals to 9.09 µs. (c). dt equals to

6.06 µs.

Fig. 4.11 shows another jetting process with almost the same conditions as that

in Fig. 4.10, but for a 30% aqueous glycerin (w/w) solution. It shows that the

cavity jet occurs much earlier for the less viscous solution. As a result, the

cavity jet pierced the balloon-like film before it becomes thin enough to

rupture on its own. For comparison, the similar jetting process for a much

higher viscosity liquid of 85% aqueous glycerin (w/w) solution is shown in

Fig. 4.12. Here the jet passed through the viscous film, without piercing it.


120 of 176

Fig. 4.11: Images showing the cavity jet pierces the thin liquid film. is 150 µm.

The liquid used is 30% aqueous glycerin (w/w) solution. Printing parameters: bi-polar

piezo-driving signal with 750 µs tdwell and 750 µs techo; driving pulse amplitude equals

to 140 V. The negative pressure inside the reservoir is -2.3 kPa relative to the


temperature is 25 ˚C. The scale bar is 200 µm. (a). dt equals to 6.06 µs. (b). dt equals

to 3.03 µs.


121 of 176

Fig. 4.12: Images showing the cavity jet fails to pierces the cavity. is 150 µm.

The liquid used is 85% aqueous glycerin (w/w) solution. Printing parameters: bi-polar

piezo-driving signal with 650 µs tdwell and 650 µs techo; driving pulse amplitude equals

to 140 V. The negative pressure inside the reservoir is -2.3 kPa relative to the


temperature is 25 ˚C. The scale bar is 200 µm. (a). dt equals to 18.18 µs. (b). dt equals

to 3.03 µs. (c). dt equals to 15.15 µs.

Fig. 4.13 shows a totally different kind of jet. It seems that two bubbles broke

up during the printing cycle, and the most interesting thing was that, there

emerged a long liquid thread between these two bubbles during the jetting. It

is also remarkable that following the bursting of the first bubble, an upstream

pressure wave was generated, which slightly deformed the lower head of the

liquid thread (as marked by the arrow in Fig. 4.13(a)).


122 of 176

Fig. 4.13: A thin liquid thread generated during the jetting. is 150 µm. The

liquid used is 70% aqueous glycerin (w/w) solution. Bi-polar piezo-driving signal

with 550 µs tdwell and 550 µs techo; 140 V. The negative pressure inside the reservoir is

-2.3 kPa. Images were taken at a frame rate of 330 kfps. Ambient temperature is 25

˚C. The scale bar is 200 µm. (a). dt equals to 3.03 µs. (b). dt equals to 12.12 µs. (c). dt

equals to 6.06 µs.

It is obviously interesting to find out how this liquid thread was produced.

However, the initial stage is missing here in Fig. 4.13, because the ultrahigh-

speed video camera [176] used to capture the jetting process can only take 103

successive frames each time. Thus we captured another similar jetting process

with a slightly earlier camera trigger and the result is shown in Fig. 4.14. As

can be seen from frame 5 in Fig. 4.14(a), when the piezo-generated cavity gets

sucked into the nozzle, a bulk of liquid was captured between it and the

preexisting bubble. When the piezo-element contracted, the preexisting bubble

pushed this bulk of liquid, deforming the bottom of the piezo-generated cavity,

and finally expelled the liquid out of the nozzle, forming a long liquid thread.


123 of 176

Fig. 4.14: Images showing the interaction between the piezo-generated cavity and the

preexisting bubble inside the nozzle. is 150 µm. The liquid used is 75% aqueous

glycerin (w/w) solution. Printing parameters: bi-polar piezo-driving signal with 550

µs tdwell and 550 µs techo; driving pulse amplitude equals to 140 V. The negative

pressure inside the reservoir is -2.3 kPa relative to the atmospheric pressure. Images

were taken at a frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is

500 µm. dt equals to 6.06 µs.


124 of 176

4.3.3 More on Surfaces Collapse Jets

Fig. 4.15 shows another type of “surfaces collapse” jet which occurs during

inkjet printing. Initially an air bubble stays deep inside the nozzle. When the

piezo-generated positive acoustic pressures reach the bubble, the focusing

flow squeezes the bubble and causes part of the air and the liquid being

expelled out of the nozzle. The squeezed air blows the liquid to an elongated

crown shape and breaks it to release part of the air. Thus the pressure inside

the cavity decreases rapidly. After only around 50 µs, the conical wall of the

broken crown contracts on the lower sections and closes up again. Due to

inertia, the expelled liquid, which constituting of the crown wall, will travel

downward along the crown wall and accumulates at the vertex of the crown.

Meanwhile, the piezo-element expands to its equilibrium state, causing a

negative pressure inside the nozzle. Thus the atmosphere pressure squeezes

the air inside the crown into the nozzle, as marked by the arrows in Fig.

4.15(b). The combination of these two effects causes a “necking” effect on the

crown, with a decreasing in the crown height and a decreasing in the angle at

the crown vertex, forming a cusp shown in closeup. The arrow in the first

frame of Fig. 4.15(c) marks this angle at the crown vertex. Finally, the crown

wall meets and collapses with itself, sending an upward fine jet into the nozzle,

as can be seen clearly in frame 4 of Fig. 4.15(c). Furthermore, a tiny air bubble

is also trapped by the collapse (frame 5 in Fig. 4.15(c)). Fig. 4.16 provides a

close view of a similar kind of surfaces collapse jet. It clearly shows that the

jets are produced when the hyperbolic surface collapses. Two jets are

generated, one upward and one downward, as marked by the arrows.


125 of 176

Fig. 4.15: Surfaces collapse jet upward into the nozzle. is 150 µm. 85% water

glycerin (w/w) solution. Bi-polar piezo signal: 650 µs tdwell and 650 µs techo; 140 V.

The negative pressure inside the reservoir is -2.3 kPa. Images were taken at 165 kfps.

Ambient temperature is 25 ˚C. (a). The scale bar is 1 mm. dt equals to 24.24 µs. (b).

The scale bar is 1 mm. dt equals to 18.18 µs. (c). The scale bar is 500 µm. dt equals to

12.12 µs.


126 of 176

Fig. 4.16: Surfaces collapse jets. is 150 µm. The liquid used is 50% aqueous

glycerin (w/w) solution. Printing parameters: bi-polar piezo-driving signal with 550

µs tdwell and 550 µs techo; driving pulse amplitude equals to 140 V. The negative

pressure inside the reservoir is -2.3 kPa relative to the atmospheric pressure. Images

were taken at a frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is

200 µm. Image for frame number n = 1, 3, 5 …… 13, 15.

4.3.4 Viscosity Effects on Jet Velocity

Fig. 4.17: Jetting velocities obtained for different concentration of aqueous glycerin

solutions (w/w): 0%, 10%, 30%, 50%, 70%, 75%, 80%, and 85%.


127 of 176

For each concentration of the aqueous glycerin solution, the printing

parameters were varied to obtain different jetting velocities as well as jet

diameters. Fig. 4.17 shows the velocities obtained during the experiments. As

can be seen, for the typical jet (Jet I), the highest jetting velocity for different

solutions fairly decreases monotonously with the increase of viscosity. While

for Jet II, the highest velocity occurs for intermediate, with a maximum peak

when 50% glycerin-water solution is printed.

Fig. 4.18 shows the fastest jet generated in the experiment. The jetting process

is similar to the one shown in Fig. 4.10(d). A piezo-generated cavity was

sucked into the nozzle and got into contact with a preexisting bubble, without

coalescing with it. The later piezo-generated radial focusing flow pushed out

the lower cavity as well as the lower cap of the preexisting bubble, forming

the balloon-like bubble. This balloon-like bubble collapsed later, followed by

the cavity jet shown here.

From extensive experiment results, it became clear that the jets produced in

this way normally have much higher velocities. Also from Fig. 4.17, it seems

that a solution with a moderate viscosity is optimal to produce a fast jet. The

reason may be due to the fact that for solutions with lower viscosities, the two

cavities inside the nozzle are more likely to coalesce (as shown in Fig. 4.6);

while for solutions with much higher viscosities, it becomes more difficult for

the viscous balloon-like bubble (as shown in Fig. 4.12) to rupture. The

bubble’s oscillation dissipates much energy; the viscous film itself exerts flow

resistance to the jet, and finally this two effects lead to a slower jet.


128 of 176

Fig. 4.18: The fastest jet observed in the experiment: a 9 µm jet with a velocity of

about 100 m/s. is 150 µm. The liquid used is 50% aqueous glycerin (w/w)




frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 200 µm. Time

interval between frames is dt = 3.03 µs.

4.3.5 Relationship between Jet Velocity and Jet

Diameter

Fig. 4.17 shows that the jets velocities generated during the experiments vary

over a wide range; same is true for the diameters of the jets. Fig. 4.19 shows 3

jets with different diameters as well as velocities. The only difference in their

printing conditions was slight change of the negative pressure inside the

reservoir. As was mentioned before, both type I and type II jets are quite

sensitive to this back pressure. As a result, big differences were obtained in

jetting velocity. From Fig. 4.19, it appears that the jet velocity decreases with

the increasing of the jet diameter. To verify this, Fig. 4.20 combines all the

obtained jet velocities with their corresponding jet diameters into one graph.

An inverse relationship can be clearly seen from the figure.


129 of 176

Fig. 4.19: Images showing the relationship between jet velocity and jet diameter. Jets

belong to type II. is 150 µm. The liquid used is 70% aqueous glycerin (w/w)




frame rate of 330 kfps. Ambient temperature is 25 ˚C. The scale bar is 200 µm. (a). A

1 µm jet with a velocity of 66 m/s. (b). A 3 µm jet with a velocity of 51 m/s. (c). A 10

µm jet with a velocity of 15 m/s.


130 of 176

Fig. 4.20: Images showing the relationship between jet velocity and jet diameter.

Data collected for both Jet I and Jet II. is 150 µm. Liquid used is 0%, 10%, 30%,

50%, 70%, 75%, 80% and 85% aqueous glycerin (w/w) solutions. Printing

parameters: bi-polar piezo-driving signal; driving pulse amplitude equals to 140 V.

Ambient temperature is 25 ˚C.

4.4 Conclusions

The formation of fine jets during the piezoelectric drop-on-demand inkjet

printing has been investigated using ultra-high-speed video imaging. The

speed of the jet can be as high as 100 m/s, which is much higher than the

typical droplet velocity during regular inkjet printing. The generation of such

fine jets has been studied over a wide range of viscosities, using 7 different

concentrations of water-glycerin solutions, giving viscosities as high as 100

times that of water. This jetting is associated with the inertial focusing of an

airpocket which is sucked into the nozzle during the printing. This occurs for

longer expansion times for the piezo-element. Two types of jets have been


131 of 176

identified during these experiments. The relationship between the speed of the

fine-jet and its diameter has also been characterized, over a range of

viscosities.

These fine jets are very sensitive to slight variations of reservoir back-pressure,

suggesting chaotic behavior. However, the diameters of the thinnest jets are of

the order of a few microns, indicating the successful generation of smaller

droplets (or jets) to diameters of smaller than 1 % of the orifice diameter.

While in existing studies [21, 24, 25], the diameter of the dispensed droplets

can be only reduced to a maximum of 60 % of the orifice diameter.

Consequently, the study provides a possible way to improve inkjet printing

resolution without reducing nozzle diameter.

Chapter 5: Cell Printing

132 of 176

5. CELL PRINTING

5.1 Introduction

Organ printing, is defined as “a rapid prototyping computer-aided 3D printing

technology, based on using layer by layer deposition of cells and/or cell

aggregates into a 3D gel with sequential maturation of the printed construct

into perfused and vascularized living tissue or organ” [137]. It is a feasible and

fast-developing technology which aims to build implantable organs to treat

diverse diseases such as cancer, loss of tissue function, or organ failure. As has

been mentioned in section 2.4, inkjet printing is a highly suitable candidate for

organ printing. The power of inkjet printing lies in its ability to deliver

picoliter volumes of materials (solutions, polymers, gels or cell ink) at high

speed (a jetting frequency of 2-10 kHz is quite common) and accuracy (several

tens of microns) on a target interface (probably non-planar surface, for

example, an organ surface), and to deliver active substances to a developing

structure in a well defined time-and-space sequence.

To guarantee successful organ printing, plenty of fundamental requirements

must be fulfilled: cells should be accurately placed into desired patterns; cells

have to survive the shear stresses experienced during the printing, and must

keep their viability. The number of cells inside each printed droplet should

also be well controlled as “empty droplet” and cell-less droplets are

undesirable for building living structures. Printed cells should be able to


133 of 176

adhere, spread and proliferate on the substrate, which are generally different

gels.

To date, different types of cells have been printed successfully and their

viability has been verified [3-10, 103]. The most comprehensive studies was

carried out by Saunders et al. [9], who used a commercial desktop printer to

dispense human fibroblast cells, for the investigation of the relationship

between cell survivability and the inkjet printing parameters. Their study

supported previous claims [4, 5, 8, 10] that cell survivability was not

significantly affected by the printing process since cell survival rates only fell

from 98% to 94% in their case, when the excitation pulse was increased from

40 to 80 V. However, in their study, the entire printing process was carried out

using a modified commercial printer, thus limiting their experiments to a fixed

nozzle diameter (60 µm) and a small range of the drop velocities (lower than

1.0 m/s). These limitations may be of importance, because the shear stresses,

which are expected to be the main factor in the killing of cells during the

printing process, are proportional to the velocity gradients within the nozzle.

To eliminate above two limitations, a squeeze mode piezoelectric DOD inkjet

printhead was designed and fabricated in-house, as has been represented in

detail in Chapter 3. A much larger range of droplet velocities can be obtained

by the novel printhead, compared to this earlier study [9]. Furthermore, the

improved design of the printhead allows us to change the nozzles while using

the same printhead main-body, thus enabling us to investigate the effects of

varying the orifice diameters on cell survival rates. It will be shown that by


134 of 176

using a small diameter nozzle and a high excitation voltage, the printing

process generates large enough shear stresses to cause significant decrease in

the cell survival rates. In fact, shear stresses have been studied extensively to

predict the damage of animal cells suspended in various laminar or turbulent

flows. This part of our study provides quantitative estimate of these effects on

cell survivability in DOD inkjet printing.

The number of cells in each printed droplet is one important factor in

optimizing cell printing, as empty droplets may be undesirable. The

probability distribution of cell numbers to ascertain desirable mean cell

concentration in the medium has been studied, to avoid “empty droplets” and

cell-less droplets. To form cell patterns, L929 rat fibroblast cells were firstly

printed onto alginate. Alginate has been increasingly utilized in tissue

engineering to support encapsulated cells and to regulate cell function, in a

manner similar to the extracellular matrices of mammalian tissues. However,

the major limitation to its use as an extracellular matrix is that alginate does

not mediate mammalian cell adhesion. To promote cell adhesion within

alginate gel, ligands such as arginine-glycine-aspartic acid (RGD), GRGDY,

KGD and VAPG can be used. Collagen is another widely used hydrogel with a

number of advantages including biodegradability, low immunogenicity and

controllable stability. Furthermore, collagen contains cell adhesion domain

sequences such as RGD, which facilitate cell adhesion for anchorage-

dependent cell types. Therefore, in later part of the study L929 cells were

printed onto collagen to form patterns.


135 of 176

5.2 Material Preparation and Experimental

Procedure

5.2.1 Preparation of Cells, Alginate and Collagen

L929 rat fibroblasts were cultured in 1× Dulbeccos Modified Eagles Medium

(DMEM, D1152). The medium was supplemented with 10% foetal bovine

serum (FBS, Gibco) and 1% Penicillin Streptomycin. FBS is widely used in

cell culture as cell growth promoting factor. Penicillin Streptomycin is a broad

spectrum bacteriostatic and bacteriocidal, with activity against gram negative

and gram positive organisms. Cells were cultured and sub-cultured in 150 cm3

culture flasks at 37 ºC, 5% CO2 and were observed under a microscope at

intervals until they grew to a full layer in the flasks. Cells were harvested by

trypsinizing with the utilization of 0.25%, 1 mM EDTA Na (Gibco) and

washing with phosphate-buffered saline (PBS, Gibco). DMEM was mixed

with the trypsinised cell solution and transferred to 50 cm3 conical tubes which

were then centrifuged at 1500 rpm for 5 min. After centrifugation, the

supernatant liquid was removed leaving the cell pellet and fresh media was

added. The final cell solution was gently agitated using a pipette to ensure

uniform distribution as well as to disrupt cell clumps. The required cell

concentrations of solutions for experiments were quantified by using a

haemocytometer (Fisher Scientific UK, Loughborough, UK).

A 1.0% (w/v) aqueous solution of sodium alginate (A2158, Sigma-Aldrich)

was prepared by suspending the polymer in distilled water. After 6 hours of


136 of 176

stirring by a magnetic stirrer, the solution was sterilized by sterile filtration,

using 0.22 µm membrane filters.

The 3 mg/ml collagen solution, (C4243, Sigma-Aldrich) was prepared by

mixing 8 parts of chilled collagen solution with 1 part of 10× PBS. The pH of

mixture was adjusted to 7.2–7.6 using 0.1 M NaOH. The pH value was

monitored carefully using pH paper. To prevent gelation, the resulting solution

was then maintained at temperature of 4 °C until ready for use.

5.2.2 Printing Experimental Setup

Printing experiments were carried out by using the self-developed squeeze

mode piezoelectric inkjet printing system. The setup is comprised of a

compressor, a pressure regulator, a reservoir, a piezo-actuated printhead, a

piezo driver, an Arrisun-5 lamp and a Photron Fastcam SA-1 camera (high-

speed-video camera), as shown in Fig. 5.1. The present printhead design is a

great improvement over conventional pintheads, as it allows for the use of

interchangeable nozzles, for the same piezoelectric transducer. The

interchangeable nozzle design allows one to easily clean or change a clogged

or damaged nozzle. The details of the design can be referred to Chapter 3 or

Ref. [140]. Before printing, all the components that will contact with cell ink

during printing process, which include the liquid reservoir, the printhead

chamber and the interchangeable nozzle, need to be properly sterilized by

autoclaving to 121 ˚C at 15 psi (pounds per square inch) for 60 minutes. The

inkjet process is highly periodic. Fig. 5.2(a) shows the droplet formation

process in a time sequence. Drop velocity can be calculated by dividing the


137 of 176

spacing between two droplets by their time difference. Fig. 5.2(b) shows

images of a few cells inside the nozzle.

Fig. 5.1: Schematic showing the DOD setup for cell printing experiment.

Fig. 5.2: Images taken by using the high-speed-video camera. (a). Image sequence

showing the formation of a 160 µm droplet from a 119 µm nozzle, taken at a frame

rate of 8,000 fps, giving time between frames of 125 µs. Liquid used was 1.0% (w/v)

aqueous solution of sodium alginate. Drop velocity is 0.74 m/s. (b). Images showing

cell motion inside the nozzle. Nozzle opening diameter is 119 µm.


138 of 176

For the study of cell survival rates, L929 rat fibroblast cell suspensions were

printed through orifices of three different diameters (119 µm, 81 µm and 36

µm) onto well plates (Costars) which contained the live-dead assay solution.

Preparation of the live-dead assay solution will be introduced later. The

electric pulses which were used to drive the piezoelectric transducer were in

the range of 52 to 140 V. Each sample was printed for approximately 20 s

with a driving frequency of 1.5 kHz for the printhead. Prior to the printing

process, a 15 µl cell suspension was deposited with a pipette into a well plate

in the same environment as the printing system, to act as a control.

For pattern printing, either alginate or collagen served as the substrate. The

1.0% (w/v) alginate was coated onto well-plate surfaces (Costar) to form

around 100-µm-thick film. Cells were dispensed onto this film using a cell ink

which contained 0.5% (w/v) calcium chloride and had a cell concentration

about 3×106 cells per ml. The crosslinking reaction occurs once the droplets

contact the alginate film. Printed samples were immediately placed into an

incubator. One hour after printing, fresh medium was carefully added into the

well plates, covering the gel surface and protecting the cells from dehydration.

Samples were transferred into incubator again and observed under a

microscope at intervals. When collagen served as substrate, 0.3% (w/v)

collagen solution was coated onto well plate surfaces to around 2-mm-thick

films and warmed up to 37 ºC for around 1 hour for gel formation. The well

plates were then placed onto a XY motion stage and L929 cell suspensions

were printed onto the gel according to the desired pattern. Printed samples

were immediately transferred into an incubator. Fresh medium was carefully


139 of 176

added into the well plates 1 hour following the printing, preventing cells death

from dehydration. Samples were transferred into incubator again and observed

under a microscope at intervals.

5.2.3 Survivability Tests

A Live-Dead Viability/Cytotoxicity Kit (L3224, Molecular Probes, Invitrogen)

was used to assess the survivability of the cells after the printing. The frozen

vials containing the assay were thawed and centrifuged briefly before use. 20

µl of the supplied 2 mM EthD-1 solution and 5 µl of the supplied 4 mM

calcein AM solution were added into 10 ml of 1× DMEM solution and mixed

thoroughly, which gave an approximately 4 µM EthD-1 and 2 µM calcein AM

working solution.

Cells were directly dispensed into well plates which each contained 100 µl of

the assay mixture, then incubated for 30 min. For each printing condition, cells

were dispensed into 5 separate petri dishes, to study the variation of survival

rates. Controls were taken directly from the cell ink before printing and put on

a set of separate petri dishes, undergoing the same environment and procedure.

The stained samples were then partly transferred onto microscope slides and

observed under a fluorescence microscope. Six images were captured from

each perti dish for cell counting. Cells that remained alive after the printing

were stained green and the damaged cells were stained red. The numbers of

alive and dead cells for each sample were tallied with respect to that of the

control which was taken prior to the printing.


140 of 176

5.3 Results and Discussion

5.3.1 Cell Survivability Study

5.3.1.1 Cell Printing

L929 cells were printed into Petri dishes containing the live-dead assay

solution, through 3 different orifices with the diameters of 119 µm, 81 µm and

36 µm. Printing was carried out over a range of excitation pulse amplitudes

from 52 to 140 V, while the rising/falling time was kept constant at 3 µs and

the dwell time, i.e. the time duration of the excitation pulse, was kept at 70 µs.

The driving frequency was held constant at 1.5 kHz. Each sample was printed

for approximately 20 s. The concentration of the cell suspension was about 1

million cells per ml.

Fig. 5.3: Graph showing influence of excitation pulse on droplet velocity. The orifice

diameters of the nozzles used were 36, 81 and 119 µm.


141 of 176

Fig. 5.4: Graph showing influence of excitation pulse voltage on droplet diameter.

The orifice diameters of the nozzles used were 36, 81 and 119 µm.

The initial average number of cells inside each droplet is fairly independent of

the voltage used to drive the piezo-element. This was verified by observations

under the microscope done within 2 hours of the printing, i.e. before

proliferation occurs. This result is consistent with the existing study [9];

however, the average number of cells depends strongly on the nozzle/droplet

size, as discussed below.

Fig. 5.3 and Fig. 5.4 show the effect of the excitation pulse (which is imparted

to the piezoelectric actuator) on the droplet velocity and droplet diameter,

respectively. It is shown that for all of the three nozzles, drop velocity and

droplet diameter increase with the increase of excitation pulse. This increase in

droplet velocity is especially pronounced for the 36 µm nozzle, where droplet

velocity increases from 2.4 to 16.6 m/s as the driving voltage increases from

60 V to 130 V. However, for the 36 µm nozzle, small satellite droplet is


142 of 176

generated once the driving voltage exceeds 70 V. To avoid satellite formation

typical DOD inkjet printing cannot generate such high velocities as used

herein. In fact, for specific nozzle size and pulse duration, there exists a

critical pulse amplitude, above which satellite droplets are produced [63].

When a satellite droplet is generated, the drop velocity is determined based on

the main droplet. The presence of the small satellite droplets is of no direct

relevance to the survival study, but will interfere with pattern printing.

Fig. 5.5: Graph showing a 95% survival rate of L929 rat fibroblast cells stained with

Calcein AM and Ethidium homodimer-1. Printed with an excitation pulse amplitude

of 116 V, at a frequency of 1.5 kHz, with rising and falling times of 3 µs. The orifice

used was 119 µm.

5.3.1.2 Cell Survivability: Effects of the Mean Shear Rate

Cell survivability after printing was quantitatively investigated by using the

LIVE-DEAD Viability/Cytotoxicity Kit as explained above. Fig. 5.5 shows a

stained sample which has a 95% survival rate.


143 of 176

Fig. 5.6: Mean cell survival rate with respect to excitation pulse amplitude for the

samples printed through the 36 µm orifice, with excitation pulse amplitude from 60 V

to 130 V, at a frequency of 1.5 kHz, with rising and falling times of 3 µs. Error bars

show the standard error from 5 replicates.

Fig. 5.6 shows the effects of excitation pulse amplitude on the mean cell

survival rate, for the 36 µm nozzle. It is shown that the survival rate falls from

95% to 78% as the excitation pulse is increased from 60 to 130 V, and the

lowest survival rate of 76% is observed when the highest voltage is

approached. The excitation pulse amplitude represents the power for the

piezoelectric actuator to dispense the droplets and this power directly affects

the droplet velocity and thereby the shear stress in the liquid. In Fig. 5.7, the

mean cell survival rates against excitation pulse amplitude for all of the three

different orifices are drawn together to compare the effects of different orifice

sizes on cell survivability.


144 of 176

Fig. 5.7: Graph showing the mean cell survival rate against excitation pulse

amplitude. Samples printed through orifices with the diameter of 36, 81 and 119 µm,

with excitation pulse amplitude from 52 to 140 V, at frequency of 1.5 kHz, with

rising and falling times of 3 µs. Each cell survival rate data was the average value

from 5 replicates.

It shows that survival rates fall from 99% to 85% for the 119 µm nozzle and

from 96% to 85% for the 81 µm nozzle. It can be seen that for the bigger

orifices, especially the 119 µm one, the printing did not produce a significant

reduction in cell survivability as the excitation pulse amplitude is increased.

This may be due to the fact that the cells used here were much smaller than the

two bigger orifices. The round-shaped L929 rat fibroblast cells are measured

to have a diameter of approximately 20 µm. It is known that shear stress in a

Newtonian fluid is proportioned to the velocity gradient in radial direction,

thus the highest shear stress is always generated in the region near the wall

during droplet dispensing. For the larger nozzles the fraction of cells moving

next to the wall is reduced, on average the cells will therefore experience less

shear stresses, which would ultimately lead to a higher survival rate.


145 of 176

Fig. 5.8: Graph showing percentage of cell death against the mean shear rate.

Samples printed through orifices with the diameter of 36 µm, 81 µm and 119 µm.

Each cell death rate data was the average value from 5 replicates.

Comparing the 3 different trendlines in Fig. 5.7, we conclude that it is not the

strength of the electric field which directly affects cell survival rate; rather it is

the fluid shear stress. Due to the highly transient nature of the flow driven

through the nozzle (see Fig. 5.2(a)), the detailed knowledge of the velocity

profile within the droplet at the nozzle tip is lacked here. Thus the mean shear

rate, which can be estimated by , where v and r are the drop velocity

and the nozzle radius, respectively, is used as a substitute for effective shear

stresses. Fig. 5.8 shows the percentage of cells that died against the mean

shear rate during the printing. It is shown that the cell death rate increases

approximately from 5% to 24% as the mean shear rate increases from 1.4×104

s-1

to 9.2×105 s

-1. The trend is more evident for the last eight data points which

correspond to the results for the 36 µm orifice. The results clearly show that

cell death does occur during the printing, especially under the effects of high

shear rates, above 5×105 s

-1.


146 of 176

One can roughly estimate the displacement thickness of the boundary layer δ

using the time duration of the piezo-signal T and the kinematic viscosity of the

liquid ν, with the well-known approximation [181], . The total

duration of the signal is T = 76 s and the viscosity of the cell ink is very

similar to that of water, i.e. ν = 10-3

m2/s, which gives δ = 15 m. The two

boundary layers therefore span 30 m, which is close to the diameter of the

smallest nozzle. The large velocity gradients inside the boundary layers are

therefore likely to submit many of the cells to the high shear stresses. The

geometry of the converging nozzle will certainly affect the true thickness of

these boundary layers, but this simple calculation suggests that their size

becomes quite significant for the smallest nozzle diameter of 36 m.

5.3.2 The Number of Cells in Each Droplet

Having investigated cell viability from the inkjet printing, the next step in

optimizing the use of such printing in tissue engineering is to uniformly

position cells in desired configurations.

Fig. 5.9 shows sections of two adjacent straight lines printed with a 60 µm

nozzle. The space between the lines is around 30 µm. There are between 1 to 5

cells observed in each droplet. This large deviation in cell numbers highlights

the random distribution of the cells inside the medium when it reaches the

nozzle, from which the droplets are dispensed. Therefore, a large average cell

concentration will be needed in the suspension to guarantee at least one cell

per droplet, with a certain high probability. To investigate the associated


147 of 176

probabilities, we performed a set of separate experiments described in what

follows.

Fig. 5.9: Droplets printed onto a dry substrate from a suspension with a concentration

of 2×106 cells per ml. Each droplet contains 1 to 5 cells. The orifice diameter of the

nozzle used was 60 µm.

The number of cells in each droplet can be thought of as a random variable,

whose distribution can then be estimated using basic probability theory [7].

The distribution of cells in the original medium is assumed to be random with

a uniform probability density. With this assumption the printing simply

represents random sampling of the liquid volume in the reservoir, with a

sphere of the same volume as the droplet . The aim is to determine the

probability that a certain number of cells are present in this volume. This can

be formulated in terms of a Bernoulli sequence of trials [182]. Each trial

consists of randomly assigning the position of the center of one cell inside the

whole liquid volume of the media . Successful trial occurs when the cell

lands inside the specified droplet. The probability of success in each trial is

therefore very small, i.e.

(5.1)


148 of 176

with a correspondingly very high probability of failure – . We do

however have a very large number of cells, i.e. a very large number of trials

are performed. Elementary probability theory shows that the probability of

getting k cells into a specific droplet in n trials becomes

(5.2)

where are the binomial coefficients. In our treatment n is a very large

number, i.e. essentially the total number of cells. In other words, ,

where N is the average cell concentration per unit volume. The above eq. 5.2

is therefore quite difficult to evaluate. However, the Poisson theorem can be

utilized to simplify this calculation, which gives

(5.3)

where the product now corresponds to the average number of cells

per droplet volume, which is denote by .

Fig. 5.10 shows this probability density function for a few different values of

, highlighting the variability of the number of cells in different droplets. As

the concentration increases the most likely number of cells shifts to larger k,

while the distribution also widens. The figure shows clearly, that there is a

finite probability of producing droplets containing no cells. The above eq. 5.3

shows that the probability of “empty droplets” is

(5.4)

which reduces exponentially with higher cell concentration in the medium.

Using this formula, it can be seen that to guarantee, with a 99% probability,

that each droplet contains at least one cell the average cell density in the


149 of 176

solution must be above cells/drop-volume.

Fig. 5.10: Graph showing the probability density distribution of the number of cells

in each droplet. For a range of different average cell concentration in the cell medium,

from dN = 0.5, 1.0, 1.5 … 3.0 cells per droplet.

To investigate the validity of our assumptions we carried out a set of

experiments where thousands of droplets were printed onto dry Petri dishes.

The diameter of the nozzle used was 130 µm, giving droplet diameter of 170

µm. The resulting diameter of the dry droplet residue was 268 µm, indicating

a spreading factor of around 1.6. Therefore the average height of the liquid

film was 46 µm. The drops dried out within about 1 min, leaving the dead

cells (as compared to the living cells in Fig. 5.11, which adhere to the

substrate and extend long filopodia) encased inside the remaining residue of

dried medium, as is shown in Fig. 5.12(a). The number of cells inside each

drop was then counted under the microscope.


150 of 176

Fig. 5.11: Optical micrographs of L929 rat fibroblast cells after 5 days in culture

following printing. Cell division can be observed (indicated by green circle)

apparently.

Fig. 5.12: Images of printed cells. (a). Cells inside dried droplet residues. The scale

bar is 50 µm. (b). Schematic showing the measurement of the radial location of each

cell, away from the center of the dried droplet residue.

Fig. 5.13 compares the distribution of the number of cells inside each droplet

with the theory in eq. 5.3. The theory shows excellent agreement except that

we observe slightly fewer empty drops that predicted and slightly more

droplets containing only one cell. The theory shows perfect agreement for

. Keep in mind that there are no free parameters in this relationship, with

the mean cell concentration coming directly from the experimental


151 of 176

results. This excellent agreement with the theory is expected to become even

better when the average number of cells in the medium increases.

Fig. 5.13: Graph showing the probability density distribution of the number of cells

in each droplet. The (□) stands for the experimental results and (--+--) stands for the

values calculated from eq. 5.3. Determined from microscope counting of cells in 800

droplets, which were dispensed within the first 4 minutes.

5.3.3 The Location of Cells inside Each Droplet

The spatial distribution of the cells within the dried drop was also studied. Fig.

5.12(b) shows how the radial location of each cell, away from the center of the

dried droplet residue, was measured. It is firstly verified that the horizontal

motion of the substrate, during the printing, does not move the cells towards a

specific direction. This might be introduced by the effective angle of impact of

the droplet, which is always less than 6˚ from the vertical. The evaporation of

the liquid during drying, could also introduce capillary-driven motions of the

cells to the edge of the drop, as is well-known from the everyday experience

of coffee stains [33, 34]. This was not observed in the resulting distribution of


152 of 176

cells in the dried spot, which is normalized by the area.

Fig. 5.14: Graph showing the probability of cell location within the dried droplet

splatter. The “radius” is the distance from the center of the cell to the center of the

dried droplet. The “Radius” is the radius of the dried droplet. “Rcell” is the radius of

the round-shaped L929 rat fibroblast cells, which has a value of approximately 10 µm.

Fig. 5.14 shows that the cells are most likely to be located near the center, with

clear reduction in cell numbers near the edge. This might be explained if the

thin lamella of liquid which is generated by the impact and precedes the

spreading, is of similar thickness as the cells. This is likely to occur in our

setup, as the Reynolds number of the impacts

, suggesting a

weak lamella traveling along the substrate. Here and are respectively the

liquid density and dynamic viscosity. and D are the impact velocity and

droplet diameter.


153 of 176

Fig. 5.15: Graph showing the average number of cells per droplet vs. time from start

of printing. Printing was carried out continuously over a period of 2.5 hours, at 120

Hz driving frequency.

Fig. 5.15 shows the long-time evolution of the average number of cells in each

droplet. The cell number is fairly uniform for the first hour and then reduces at

approximately a uniform rate, which is probably due to slow coagulating or

settling of the cells in the liquid chamber.


154 of 176

5.3.4 Printing Patterns

Fig. 5.16: Image showing cells printed onto a dry Petri-dish, forming an “NUS”

pattern. Each droplet contains 2 to 6 cells. The orifice diameter of the nozzle used

was 60 µm.

Fig. 5.17: Image showing a continuous line of overlapping droplets with around 6 to

8 cells per droplet in the crosslinked gel. The orifice diameter of the nozzle used was

60 µm.

For the printing of cell patterns, a manual micro-meter x-y-stage is used, for a

proof-of-concept demonstration. Fig. 5.16 shows cells fired onto a dry Petri-

dish through a 60 µm orifice, forming an “NUS” pattern. Using the manually

operated stage and single-drop printing the formation of this entire pattern

took about 5 min. As a result, most of the droplets dried up during the printing,

leaving only dead cells (due to dehydration) inside the dried outline of the

droplets. A viable substrate is necessary in order to maintain suitable moisture


155 of 176

to prevent cell death from dehydration. Fig. 5.17 shows the printed cells inside

the crosslinked gel made from 1.0% (w/v) alginate and 0.5% (w/v) calcium

chloride. (Alginate was coated onto a well-plate before printing, while calcium

chloride was mixed within the cell ink). The overlapping droplets form a

continuous straight edged line. It was subsequently found that fibroblast cells

retained their spherical shape rather than extending filopodia, which meant

that the cells failed to adhere to the alginate. The same result has also been

reported by Kuo et al. [183].

Fig. 5.18: Image showing live cells printed onto a collagen gel, forming an “NUS”

pattern. The orifice diameter of the nozzle used was 60 µm. Picture taken 5 day after

printing.

In Fig. 5.18, the same “NUS” pattern was created by dispensing the cells onto

a collagen gel. Printed cells were immediately placed into an incubator. 1 hour

following the printing, fresh medium was added into the well plates. The

samples were transferred into the incubator again and observed under

microscope at intervals. After 5 days, Live/Dead assay was applied to the

samples. A bright green fluorescence was observed after incubation for 30 min.

The cells were shown to survive after printing, adhere to the gel, spread and

proliferate, forming a denser pattern. It is worth noting that the cells were


156 of 176

slightly moved from their initial position, perhaps caused by the addition of

the fresh medium, thereby slightly reducing the resolution of the printing.

5.5 Conclusions

The study has demonstrated that piezoelectric DOD inkjet printing is able to

successfully deliver L929 rat fibroblast cells through nozzles as small as 36

µm. There was no significant cell death when dispensing the cells through the

81 µm and the 119 µm nozzle, with the mean survival rates only reducing

from 98% to 85%. This is in good agreement with the study of Saunders et al.

[9], in which a commercial printer was used to print human fibroblast cells.

When the orifice was reduced to 36 µm, the corresponding cell survival rates

fell from 95% to 76% when the excitation pulse amplitude increased from 60

V to 130 V. These results indicate that the droplet ejection out of the nozzle

has exerted large shear stresses on the cells and possibly disrupted the cell

membrane and killed about 20% of the cells. Mean shear rate was estimated

by combining the effects of droplet velocity and orifice diameter and was

correlated with the cell survival rate. A large range of mean shear rates from

1.3×104 s

-1 to 9.2×10

5 s

-1 were generated and cell survival rates were found to

be strongly affected by the higher mean shear rates, especially when the shear

rate exceeds 5×105 s

-1.

The distribution of the number of cells within each droplet was also

investigated. This was done to find the minimal cell concentration in the

medium, which is required to avoid the appearance of empty droplets, since


157 of 176

droplets containing no cells may be detrimental to pattern printing. The

distribution of cell numbers is found to have a binomial form, which

consistent with a uniform distribution of cells inside the medium in the

reservoir.

For pattern printing, L929 fibroblast cells were delivered by using a 60 µm

nozzle. Printed cells successfully kept their patterns in the crosslinked gel

made from 1.0% (w/v) alginate and 0.5% (w/v) calcium chloride. However, it

was found that the cells failed to adhere to alginate. On the other hand, cells

dispensed onto collagen gel were found to successfully maintain their viability,

adhere to the gel, spread and proliferate, forming a denser pattern. However,

unlike the crosslinked calcium-alginate which can immobilize cells quite

rapidly, cell adhesion to collagen needs a relatively long time to get

established. Therefore, some of the printed cells were slightly moved from

their initial position when the sample was disturbed, by the addition of fresh

medium or unintended shaking of the sample, which will reduce the resolution

of the printing. The smallest nozzle, with orifice diameter of 36 µm, was not

used for pattern printing, due to issues concerning the reliability of the printing

process, as it can easily get clogged.

Future studies should involve experiment with more mammalian cell types. It

is also of interest to check whether adding ligands or collagen into alginate

(before the crosslinking reactions) will promote cell adhesion onto the

substrate. If this can be successfully implemented, the accuracy of the pattern

printing will be significantly improved.

Chapter 6: Recommendations For Future Work

158 of 176

6. RECOMMENDATIONS FOR FUTURE

WORK

6.1 Printhead Design

As a manufacturing tool, high speed jetting is required to increase productivity

of inkjet printing technology. For a specific printhead, its maximum jetting

frequency is mainly depended on the construction of the printhead as well as

the driving signal. Typical DOD printheads generate droplets at rates in the

range 0.1-10 kHz. While the maximum jetting frequency for our in-house-

developed printhead is only 3.6 kHz. Furthermore, our printhead has a

relatively lower threshold frequency, above which jetting frequency starts to

have great effects on droplet velocity and droplet diameter. This threshold

frequency is also mainly determined by the construction of the prinhtead. Thus

future work should include a systematic study of the relationship between this

threshold frequency and printhead construction, to optimize the printhead

design, which includes parameters such as the piezo-ceramic material, the

dimensions of the piezo-element, the dimensions of the printhead chamber, the

nozzle profile, etc..

The dispensing of relatively high concentration of sodium alginate solutions

shows that the printing behavior of non-Newtonian liquid is distinctly different

from that of Newtonian liquids. Thus more experiments need to be carried out


159 of 176

to investigate the characteristics of our printhead for printing non-Newtonian

liquids or even particle-laden liquids.

6.2 Reducing Droplet Size

The results in Chapter 3 show that by carefully control the driving signal, fine

jets can be produced with a relatively bigger nozzle size. The results show the

possibility of reducing droplet size (or improving printing resolution) without

reducing nozzle size. However, the repeatability of the method is still far from

perfect. It is interesting to find out whether the repeatability of the generation

of such fine jet will be improved by adopting different nozzle profiles.

Furthermore, current experiments are collected only from the printing of

Newtonian liquids; more experiments also need to be carried out to investigate

the generation of fine jets with non-Newtonian liquids.

6.3 Cell Printing

To fully understand the long-term effects of stress forces on cell viability,

more experiments need to be carried out to collect quantitative data for cell

adhesion, cell spreading and cell migration, and cell proliferation. It is also of

interesting to compare our cell viability results with cells in other fluidic

systems, under similar shear stress forces. Future studies could also involve

experiment with more mammalian cell types. It is also of interest to check

whether adding ligands or collagen into the alginate (before the crosslinking

reactions) will promote cell adhesion onto the substrate. If this can be


160 of 176

successfully implemented, the accuracy of the pattern printing will be

significantly improved.

Bibliography

161 of 176

Bibliography

1. Kyocera Corporation, website cited:

http://global.kyocera.com/news/2009/0903_kdos.html

2. P. Calvert: “Printing cells”, Science, Vol. 318, No. 5848, 2007, pp. 208-

209.

3. R. Chang, J. Nam, W. Sun: “Effects of dispensing pressure and nozzle

diameter on cell survival from solid freeform fabrication-based direct cell

writing”, Tissue Engineering Part A, Vol. 14, No. 1, 2008, pp. 41-48.

4. T. Xu, J. Jin, C. Gregory, J. J. Hichman, T. Boland: “Inkjet printing of

viable mammalian cells”, Biomaterials, Vol. 26, No. 1, 2005, pp. 93-99.

5. T. Xu, S. Petridou, E. H. Lee, E. A. Roth, N. R. Vyavahare, J. J. Hickman,

T. Boland: “Construction of high-density bacterial colony arrays and

patterns by the ink-jet method”, Biotechnology and Bioengineering, Vol.

85, No. 1, 2004, pp. 29-33.

6. C. Y. Chen, J. A. Barron, B. R. Ringeisen: “Cell patterning without

chemical surface modification: Cell-cell interactions between printed

bovine aortic endothelial cells (BAEC) on a homogeneous cell-adherent

hydrogel”, Applied Surface Science, Vol. 252, No. 24, 2006, pp. 8641-

8645.

7. J. A. Barron, D. B. Krizman, B. R. Ringeisen: “Laser printing of single

cells: statistical analysis, cell viability, and stress”, Annals of Biomedical

Engineering, Vol. 33, No. 2, 2005, pp. 121-130.

8. M. Nakamura, A. Kobayashi, F. Takagi, A. Watanabe, Y. Hiruma, K.

Ohuchi, Y. Iwasaki, M. Horie, I. Morita, S. Takatani: “Biocompatible

inkjet printing technique for designed seeding of individual living cells”,

Tissue Engineering, Vol. 11, No. 11-12, 2005, pp. 1658-1666.

9. R. E. Saunders, J. E. Gough, B. Derby: “Delivery of human fibroblast

cells by piezoelectric drop-on-demand inkjet printing”, Biomaterials, Vol.

29, No. 2, 2008, pp. 193-203.

10. T. Xu, C. A. Gregory, P. Molnar, X. Cui, S. Jalota, S. B. Bhaduri, T.

Boland: “Viability and electrophysiology of neural cell structures

generated by the inkjet printing method”, Biomaterials, Vol. 27, No. 19,

2006, pp. 3580-3588.

11. Cambridge Display Technology, Ltd., website cited:

http://www.tinhoahoc.com/Semiconductor/IJ-PLED1.pdf

http://global.kyocera.com/news/2009/0903_kdos.html

http://www.tinhoahoc.com/Semiconductor/IJ-PLED1.pdf

Bibliography

162 of 176

12. B. J. de Gans, P. C. Duineveld, U. S. Schubert: “Inkjet printing of

polymers: state of the art and future developments”, Advanced Materials,

Vol. 16, No. 3, 2004, pp. 203-213.

13. S. E. Molesa: “Ultra-low-cost printed electronics”, Ph.D thesis from

University of California at Berkeley, 2006.

14. D. H. Redinger: “Solution processing techniques for low-cost circuit

fabrication”, Ph.D thesis from University of California at Berkeley, 2007.

15. C. W. Sele, T. von Werne, R. H. Friend, H. Sirringhaus: “Lithography-

free, self-aligned inkjet printing with sub-hundred-nanometer resolution”,

Advanced Materials, Vol. 17, No. 8, 2005, pp. 997-1001.

16. D. J. Gundlach, T. N. Jackson, D. G. Schlom, S. F. Nelson: “Solvent-

induced phase transition in thermally evaporated pentacene films”,

Applied Physics Letters, Vol. 74, No. 22, 1999, pp. 3302-3304.

17. J. Perelaer, B. J. de Gans, U. S. Schubert: “Ink-jet printing and microwave

sintering of conductive silver tracks”, Advanced Materials, Vol. 18, No.

16, 2006, pp. 2101-2104.

18. N. Stutzmann, R. H. Friend, H. Sirringhaus: “Self-aligned, vertical-

channel, polymer field-effect transistors”, Science, Vol. 299, No. 5614,

2009, pp. 1881-1884.

19. Y. Wang: “Applying drop-on-demand inkjet printing method to maskless

lithography”, Ph.D thesis from University of California at Berkeley, 2005.

20. H. Sirringhaus, T. Kawase, R. H. Friend, T. Shimoda, M. Inbasekaran, W.

Wu, E. P. Woo: “High-resolution inkjet printing of all-polymer transistor

circuits”, Science, Vol. 290, 2000, pp. 2123-2126.

21. A. U. Chen, O. A. Basaran: “A new method for significantly reducing

drop radius without reducing nozzle radius in drop-on-demand drop

production”, Physics of Fluids, Vol. 14, No. 1, 2002, pp. L1-L4.

22. A. A. Goghari, S. Chandra: “Producing droplets smaller than the nozzle

diameter by using a pneumatic drop-on-demand droplet generator”,

Experiments in Fluids, Vol. 44, No. 1, 2008, pp. 105-114.

23. C. Rensch: “Creation of small microdrops”, MicroFab Technologies Inc.,

Oct 2006.

24. H. Y. Gan, X. Shan, T. Eriksson, B. K. Lok, Y. C. Lam: “Reduction of

droplet volume by controlling actuating waveforms in inkjet printing for

micro-pattern formation”, Journal of micromechanics and

microengineering, Vol. 19, 2009, 055010.

Bibliography

163 of 176

25. N. Riefler, T. Wriedt: “Generation of monodisperse micron-sized droplets

using free adjustable signals”, Particle & Particle Systems

Characterization, Vol. 25, No. 2, 2008, pp. 176-182.

26. H. P. Le: “Progress and Trends in Ink-jet Printing Technology”, Journal

of Imaging Science and Technology, Vol. 42, No. 1, 1998, pp. 49-62.

27. L. Rayleigh: “On the instability of jets”, Proceedings of the London

Mathematical Society, Vol. 10, No. 4, 1878, pp. 4-13.

28. J. Heinzl, C. H. Hertz: “Ink-jet printing,” Advances in Electronics and

Electron Physics, Vol. 65, 1985, pp. 91–171.

29. A. Mikalesen: “Ink jet apparatus and method employing phase change

ink”, US Patent No. 4,742,364, 1988.

30. A. Benjeddou: “Shear-mode piezoceramic advanced materials and

structures: a state of the art”, Mechanics of Advances Materials and

Structures, Vol. 14, No. 4, 2007, pp. 263-275.

31. H. R. Kang: “Water-based ink-jet ink. 3. performances studies”, Journal

of Imaging Science, Vol. 35, No. 3, 1991, pp. 195-201.

32. S. C. Chang, J. Liu, J. Bharathan, Y, Yang, J. Onohara, J. Kido:

“Multicolor Organic Light-Emitting Diodes Processed by Hybrid Inkjet

Printing”, Advanced Materials, Vol. 11, No. 9, 1999, pp. 734-737.

33. M. Singh, H. M. Haverinen, P. Dhagat, G. E. Jabbour: “Inkjet printing -

progress and its applications”, Advanced Materials, Vol. 22, 2010, pp.

673-685.

34. R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nagel, T. A.

Witten: “Capillary flow as the cause of ring stains from dried liquid

drops”, Nature, Vol. 389, 1997, pp. 827-829.

35. M. Ikegawa, H. Azuma: “Droplet behaviors on substrates in thin-film

formation using ink-jet printing”, JSME International Journal Series B:

Fluids and Thermal Engineering, Vol. 47, No. 3, 2004, pp. 490-496.

36. Xaar, website cited: http://www.xaar.com/printheads.aspx

37. MicroFab Technologies, Inc., website cited:

http://www.microfab.com/equipment/systems.html

38. M. A. Groninger, P. G. M. Kruijt, H. Reinten, R. H. Schippers, J. M. M.

Simons: “Method of controlling an inkjet printhead, an inkjet printhead

suitable for use of said method, and an inkjet printer comprising said

printhead”, US Patent 7,357,474, 2008.

http://www.xaar.com/printheads.aspx

http://www.microfab.com/equipment/systems.html

http://www.wikipatents.com/as/s_inventor/Groninger%3B+Mark+Alexander

http://www.wikipatents.com/as/s_inventor/Kruijt%3B+Pieter+Gijsbertus+Maria

http://www.wikipatents.com/as/s_inventor/Reinten%3B+Hans

http://www.wikipatents.com/as/s_inventor/Schippers%3B+Ronald+Herman

http://www.wikipatents.com/as/s_inventor/Simons%3B+Johannes+Mathieu+Marie

Bibliography

164 of 176

39. K. S. Kwon: “Waveform design methods for piezo inkjet dispensers based

on measured meniscus motion”, Journal of Microelectromechanical

Systems, Vol. 18, No. 5, 2009, pp. 1118-1125.

40. K. S. Kwon, W. Kim: “A waveform design method for high-speed inkjet

printing based on self-sensing measurement”, Sensors and Actuators: A,

Vol. 140, 2007, pp. 75-83.

41. S. Bruner, D. Xu, C. Phillips: “Drop landing accuracy improvements in

inkjet printed OLED displays”, SID Symposium Digest of Technical

Papers, Vol. 38, 2007, pp. 1611-1612.

42. M. McDonald: “Manufacture of flat panel displays using piezoelectric

drop-on-demand ink jet”, SID Symposium Digest of Technical Papers,

Vol. 34, 2003, pp. 1186-1189.

43. T. Shimoda, K. Morii, S. Seki, H. Kiguchi: “Inkjet printing of light-

emitting polymer displays”, MRS Bulletin, Vol. 28, 2003, pp. 821-827.

44. R. A. Street, W. S. Wong, S. E. Ready, M. L. Chabinyc, A. C. Arias, S.

Limb, A. Salleo, R. Lujan: “Jet printing flexible displays”, Materials

Today, Vol. 9, No. 4, 2006, pp. 32-37.

45. J. X. Zhou, J. Y. H. Fuh, H. T. Loh, Y. S. Wong, Y. S. Ng, J. G. Jeffrey, S.

J. Chua: “Characterization of drop-on-demand microdroplet printing”,

The International Journal of Advanced Manufacturing Technology, Vol.

48, 2010, pp. 243-250.

46. D. Soltman, V. Subramanian: “Inkjet-printed line morphologies and

temperature control of the coffee ring effect”, Langmuir, Vol. 24, No. 5,

2008, pp. 2224-2231.

47. P. J. Smith, D. Y. Shin, J. E. Stringer, B. Derby, N. Reis: “Direct ink-jet

printing and low temperature conversion of conductive silver patterns”,

Journal of Materials Science, Vol. 41, 2006, pp. 4153-4158.

48. J. F. Dijksman: “Hydrodynamics of small tubular pumps”, Journal of

Fluid Mechanics, Vol. 139, 1984, pp. 173-191.

49. J. E. Fromm: “Numerical calculation of the fluid dynamics of drop-on-

demand jets”, IBM Journal of Research and Development, Vol. 28, No. 3,

1984, pp. 322-333.

50. E. B. Wylie, V. L. Streeter: “Fluid Transients”, New York: McGraw-Hill,

1978.

51. N. Reis, C. Ainsley, B. Derby: “Ink-jet delivery of particle suspensions by

piezoelectric droplet ejectors”, Journal of Applied Physics, Vol. 97, 2005,

094903.

Bibliography

165 of 176

52. D. B. Bogy, F. E. Talke: “Experimental and theoretical study of wave

propagation phenomena in drop-on-demand ink jet devices”, IBM Journal

of Research and Development, Vol. 28, No. 3, 1984, pp. 314-321.

53. H. Wijshoff: “Free surface flow and acousto-elastic interaction in piezo

inkjet”, Technical Proceedings of the 2004 NSTI Nanotechnology

Conference and Trade Show, Vol. 2, 2004, 215-218.

54. H. C. Wu, H. J. Lin, W. S. Hwang: “A numerical study of the effect of

operating parameters on drop formation in a squeeze mode inkjet device”,

Modelling and Simulation in Materials Science and Engineering, Vol. 13,

2005, pp. 17-34.

55. H. C. Wu, T. R. Shan, W. S. Hwang, H. J. Lin: “Study of micro-droplet

behavior for a piezoelectric inkjet printing device using a single pulse

voltage pattern”, Materials Transactions, Vol. 45, No. 5, 2004, pp. 1794-

1801.

56. R. Li, N. Ashgriz, S. Chandra: “Droplet generation from pulsed micro-

jets”, Experimental Thermal and Fluid Science, Vol. 32, No. 8, 2008, pp.

1679-1686.

57. A. S. Yang, W. M. Tsai: “Ejection process simulation for a piezoelectric

microdroplet generator”, Transactions of the American Society of

Mechanical Engineers, Vol. 128, No. 6, 2006, pp. 1144-1152.

58. Y. Wang, J. Bokor, A. Lee: “Maskless lithography using drop-on-demand

inkjet printing method”, Proceedings of SPIE, Vol. 5374, 2004, pp. 628-

636.

59. B. J. de Gans, E. Kazancioglu, W. Meyer, U. S. Schubert: “Ink-jet

printing polymers and polymer libraries using micropipettes”,

Macromolecular Rapid Communications, Vol. 25, No. 1, 2004, pp. 292-

296.

60. Microfab Technote 99-03: “Drive waveform effects on ink-jet device

performance”, 1999.

61. S. L. Zoltan: “Pulse droplet ejection system”, US Patent 3,683,212, 1974.

62. K. A. M. Seerden, N. Reis, J. R. G. Evans, P. S. Grant, J. W. Halloran, B.

Derby: “Ink-jet printing of wax-based alumina suspensions”, Journal of

The American Ceramic Society, Vol. 84, No. 11, 2001, pp. 2514-2520.

63. H. Dong, W. W. Carr, J. F. Morris: “An experimental study of drop-on-

demand drop formation”, Physics of Fluids, Vol. 18, 2006, 072102.

64. H. Dong, W. W. Carr, J. F. Morris: “Visualization of drop-on-demand

inkjet: Drop formation and deposition”, Review of Scientific Instruments,

Vol. 77, 2006, 085101.

Bibliography

166 of 176

65. H. J. Shore, G. M. Harrison: “The effect of added polymers on the

formation of drops ejected from a nozzle”, Physics of Fluids, Vol. 17,

2005, 033104.

66. F. G. Tseng, C. J. Kim, C. M. Ho: “A high-resolution high-frequency

monolithic top-shooting microinjector free of satellite drops-Part I:

concept, design, and model”, Journal of Microelectromechanical Systems,

Vol. 11, No. 5, 2002, pp. 427-436.

67. Boston Piezo-Optics Inc, website cited:

http://bostonpiezooptics.com/?D=6

68. E. R. Lee: “Microdrop generation”, London: CRC Press, 2003.

69. D. J. Hayes, D. B. Wallace: “Overview of small holes”, Non Traditional

Machining Conference, MS89184, Orlando, Florida, November 1989.

70. H. Robert: “Properties of crystalline silicon”, London: The Institution of

Electrical Engineers, 1999.

71. E. Bassous, H. H. Taub, L. Kuhn: “Ink jet printing nozzle arrays etched in

silicon”, Applied Physics Letters, Vol. 31, No. 2, 1977, pp. 135-137.

72. L. Kuhn, E. Bassous, R. Lane: “Silicon charge electrode array for ink jet

printing”, IEEE Transactions on Electron Devices, Vol. ED-25, No. 10,

1978, pp. 1257-1260.

73. T. Laurell, L. Wallman, J. Nilsson: “Design and development of a silicon

microfabricated flow-through dispenser for on-line picolitre sample

handling”, Journal of Micromechanics and Micro engineering, Vol. 9,

1999, pp. 369-376.

74. R. F. Service: “Printable electronics that stick around”, Science, Vol. 304,

No. 5671, 2004, pp. 675.

75. H. E. Katz: “Recent advances in semiconductor performance and printing

processes for organic transistor-based electronics”, Chemistry of

Materials, Vol. 16, No. 23, 2004, pp. 4748-4756.

76. J. A. Rogers, Z. Bao: “Printed plastic electronics and paperlike displays”,

Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 40, No. 20,

2002, pp. 3327-3334.

77. J. A. Rogers, Z. Bao, K. Baldwin, A. Dodabalapur, B. Crone, V. R. Raju,

V. Kuck, H. Katz, K. Amundson, J. Ewing, P. Drzaic: “Paper-like

electronic displays: large-area rubber-stamped plastic sheets of electronics

and microencapsulated electrophoretic inks”, Proceedings of the National

Academy of Science of the United States of America, Vol. 98, No. 9,

2001, pp. 4835-4840.

http://bostonpiezooptics.com/?D=6

Bibliography

167 of 176

78. T. Sekitani, Y. Noguchi, U. Zschieschang, H. Klauk, T. Someya:

“Organic transistors manufactured using inkjet technology with

subfemtoliter accuracy”, Proceedings of the National Academy of

Sciences of the United States of America, Vol. 105, No. 13, 2008, pp.

4976-4980.

79. N. Zhao, M. Chiesa, H. Sirringhaus, Y. Li, Y. Wu, B. Ong: “Self-aligned

inkjet printing of highly conducting gold electrodes with submicron

resolution”, Journal of Applied Physics, Vol. 101, No. 6, 2007, 064513.

80. J. Z. Wang, J. Gu, F. Zenhausern, H. Sirringhaus: “Low-cost fabrication

of submicron all polymer field effect transistors”, Applied Physics Letters,

Vol. 88, No. 13, 2006, 133502.

81. J. Z. Wang, Z. H. Zheng, H. W. Li, W. T. S. Huck, H. Sirringhaus:

“Dewetting of conducting polymer inkjet droplets on patterned surfaces”,

Nature Materials, Vol. 3, No. 3, 2004, pp. 171-176.

82. M. Caironi, E. Gili, T. Sakanoue, X. Cheng, H. Sirringhaus: “High yield,

single droplet electrode arrays for nanoscale printed electronics”, Acs

Nano, Vol. 4, No. 3, 2010, pp. 1451-1456.

83. B. J. de Gans, S. Hoeppener, U. S. Schubert: “Polymer relief

microstructures by inkjet etching”, Journal of Materials Chemistry, Vol.

17, 2007, pp. 3045-3050.

84. C. Y. Kung, M. D. Barnes, N. Lermer, W. B. Whitten, J. M. Ramsey:

“Single-molecule analysis of ultradilute solutions with guided streams of

1-µm water droplets”, Applied Optics, Vol. 38, No. 9, 1999, pp. 1481-

1487.

85. S. Temple: “Small fast inkdrop emission from a nozzle”, 43rd

European

Study Group with Industry Technical Report, Lancaster University, UK

2002.

86. J. U. Park, M. Hardy, S. J. Kang, K. Barton, K. Adair, D. K.

Mukhopadhyay, C. Y. Lee, M. S. Strano, A. G. Alleyne, J. G. Georgiadis,

P. M. Ferreira, J. A. Rogers: “High-resolution electrohydrodynamic jet

printing”, Nature Materials, Vol. 6, No. 10, 2007, pp. 782-789.

87. S. N. Jayasinghe, M. J. Edirisinghe: “Electric-field driven jetting from

dielectric liquids”, Applied Physics Letters, Vol. 85, No. 18, 2004, pp.

4243-4245.

88. S. N. Jayasinghe, A. N. Qureshi, P. A. M. Eagles: “Electrohydrodynamic

jet processing: an advanced electric-field-driven jetting phenomenon for

processing living cells”, Small, Vol. 2, No. 2, 2006, pp. 216-219.

http://apps.isiknowledge.com.libproxy1.nus.edu.sg/full_record.do?product=WOS&search_mode=GeneralSearch&qid=3&SID=Y1b87dnnLFe3bPFB1C5&page=1&doc=3

http://apps.isiknowledge.com.libproxy1.nus.edu.sg/full_record.do?product=WOS&search_mode=GeneralSearch&qid=3&SID=Y1b87dnnLFe3bPFB1C5&page=1&doc=3

Bibliography

168 of 176

89. I. Marginean, L. Parvin, L. Heffernan, A. Vertes: “Flexing the electrified

meniscus: the birth of a jet in electrosprays”, Analytical Chemistry, Vol.

76, No. 14, 2004, pp. 4202-4207.

90. G. Taylor: “Disintegration of water drops in an electric field”,

Proceedings of the Royal Society of London. Series A, Mathematical and

Physical Sciences, Vol. 280, No. 1382, 1964, pp. 383-397.

91. J. F. de la Mora: “The fluid dynamics of Taylor cones”, Annual Review of

Fluid Mechanics, Vol. 39, 2007, pp. 217-243.

92. H. K. Choi, J. U. Park, O. O. Park, P. M. Ferreira, J. G. Georgiadis, J. A.

Rogers: “Scaling laws for jet pulsations associated with high-resolution

electrohydrodynamic printing”, Applied Physics Letters, Vol. 92, No. 12,

2008, 123109.

93. K. Murata, J. Matsumoto, A. Tezuka, Y. Matsuba, H. Yokoyama: “Super-

fine ink-jet printing: toward the minimal manufacturing system”,

Microsystem Technologies-Micro-and Nanosystems-Information Storage

and Processing Systems, Vol. 12, 2005, pp. 2-7.

94. C. H. Chen, D. A. Saville, I. A. Aksay: “Scaling laws for pulsed

electrohydrodynamic drop formation”, Applied Physics Letters, Vol. 89,

No. 12, 2006, 124103.

95. J. L. Li: “On the meniscus deformation when the pulsed voltage is

applied”, Journal of Electrostatics, Vol. 64, No. 1, 2006, pp. 44-52.

96. J. L. Li: “EHD sprayings induced by the pulsed voltage superimposed to a

bias voltage”, Journal of Electrostatics, Vol. 65, No. 12, 2007, pp. 750-

757.

97. J. L. Li: “On the stability of electrohydrodynamic spraying in the cone-jet

mode”, Journal of Electrostatics, Vol. 65, No. 4, 2007, pp. 251-255.

98. W. Balachandran, W. Machowski, C. N. Ahmad: “Electrostatic

atomization of conducting liquids using AC superimposed on DC fields”,

IEEE Transactions on Industry Applications, Vol. 30, No. 4, 1994, pp.

850-855.

99. Y. J. Kim, J. S. Lee, S. Y. Kim, S. E. Park, J. Hwang, Y. J. Kim: “Hybrid

on demand jetting system for ultra fine droplet based on

electrohydrodynamic and piezoelectric actuation”, IEEE 22nd

International Conference on Micro Electro Mechanical Systems, 2009, pp.

491-494.

100. Y. J. Kim, S. Y. Kim, J. S. Lee, J. Hwang, Y. J. Kim: “On-demand

electrohydrodynamic jetting with meniscus control by a piezoelectric

actuator for ultra-fine patterns”, Journal of Micromechanics and

Microengineering, Vol. 19, No. 10, 2009, 107001.

Bibliography

169 of 176

101. H. Yamazoe, T. Tanabe: “Cell micropatterning on an albumin-based

substrate using an inkjet printing technique”, Journal of Biomedical

Materials Research: Part A, Vol. 91A, No. 4, 2009, pp. 1202-1209.

102. Y. Ito: “Surface micropatterning to regulate cell functions”, Biomaterials,

Vol. 20, No. 23-24, 1999, pp. 2333-2342.

103. S. Takahashi, Y. Tomikawa: “A study of handling of living mammalian

cells using a piezoelectric dispensing head”, Japanese Journal of Applied

Physics, Vol. 42, 2003, pp. 3098-3101.

104. Y. Chisti: “Hydrodynamic damage to animal cells”, Critical Reviews in

Biotechnology, Vol. 21, No. 2, 2001, pp. 67-110.

105. C. Born, Z. Zhang, M. Alrubeai, C. R. Thomas: “Estimation of disruption

of animal cells by laminar shear stress”, Biotechnology and

Bioengineering, Vol. 40, No. 9, 1992, pp. 1004-1010.

106. Z. Zhang, M. Alrubeai, C. R. Thomas: “Estimation of disruption of

animal cells by turbulent capillary flow”, Biotechnology and


107. B. J. H. Zoro, S. Owen, R. A. L. Drake, M. Hoare: “The impact of process

stress on suspended anchorage-dependent mammalian cells as an indicator

of likely challenges for regenerative medicines”, Biotechnology and


108. R. E. Saunders, J. E. Gough, B. Derby: “Inkjet printing of mammalian

primary cells for tissue engineering applications”, In: Laurencin CT,

Botchwey, EA, editors, Nanoscale Materials Science in Biology and

Medicine, Boston: MRS, 2005, pp. 57-62.

109. H. J. Kong, M. K. Smith, D. J. Mooney: “Designing alginate hydrogels to

maintain viability of immobilized cells”, Biomaterials, Vol. 24, No. 22,

2003, pp. 4023-4029.

110. L. Shapiro, S. Cohen: “Novel alginate sponges for cell culture and

transplantation”, Biomaterials, Vol. 18, No. 8, 1997, pp. 583-590.

111. G. Klock, A. Pfeffermann, C. Ryser, P. Grohn, B. Kuttler, H. J. Hahn, U.

Zimmermann: “Biocompatibility of mannuronic acid-rich alginates”,

Biomaterials, Vol. 18, No. 10, 1997, pp. 707-713.

112. E. Alsberg, K.W. Anderson, A. Albeiruti, R. T. Franceschi, D. J. Mooney:

“Cell-interactive alginate hydrogels for bone tissue engineering”, Journal

of Dental Research, Vol. 80, No. 11, 2001, pp. 2025-2029.

113. A. D. Augst, H. J. Kong, D. J. Mooney: “Alginate hydrogels as

biomaterials”, Macromolecular Bioscience, Vol. 6, No. 8, 2006, pp. 623-

633.

Bibliography

170 of 176

114. N. E. Fedorovich, J. Alblas, J. R. de Wijn, W. E. Hennink, A. J. Verbout,

W. J. A. Dhert: “Hydrogels as extracellular matrices for skeletal tissue

engineering: state-of-the-art and novel application in organ printing”,

Tissue Engineering, Vol. 13, No. 8, 2007, pp. 1905-1925.

115. B. K. Mann, A. T. Tsai, T. Scott-Burden, J. L. West: “Modification of

surfaces with cell adhesion peptides alters extracellular matrix deposition”,

Biomaterials, Vol. 20, No. 23-24, 1999, pp. 2281-2286.

116. S. P. Massia, J. A. Hubbell: “Covalent surface immobilization of Arg-

Gly-Asp- and Tyr-Ile-Glv-Ser-Arg-containing peptides to obtain well-

defined cell-adhesive substrates”, Analytical Biochemistry, Vol. 187, No.

2, 1990, pp. 292-301.

117. K. C. Olbrich, T. T. Andersen, F. A. Blumenstock, R. Bizios: “Surfaces

modified with covalently-immobilized adhesive peptides affect fibroblast

population motility”, Biomaterials, Vol. 17, No. 8, 1996, pp. 759-764.

118. J. A. Rowley, G. Madlambayan, D. J. Mooney: “Alginate hydrogels as

synthetic extracellular matrix materials”, Biomaterials, Vol. 20, No. 1,

1999, pp. 45-53.

119. M. G. Patino, M. E. Neiders, S. Andreana, B. Noble, R. E. Cohen:

“Collagen: an overview”, Implant Dentistry, Vol. 11, No. 3, 2002, pp.

280-285.

120. F. H. Silver, G. Pins: “Cell-growth on collagen-a review of tissue

engineering using scaffolds containing extracellular-matrix”, Journal of

Long-Term Effects of Medical Implants, Vol. 2, No. 1, 1992, pp. 67-80.

121. W. Sun: “Computer-aided tissue engineering, Part II”, Tutorial

Presentation at CAD/CG’05, Hong Kong University of Science and

Technology, 2005.

122. R. Landers, U. Hübner, R. Schmelzeisen, R. Mülhaupt: “Rapid

prototyping of scaffolds derived from thermoreversible hydrogels and

tailored for applications in tissue engineering”, Biomaterials, Vol. 23, No.

23, 2002, pp. 4437-4447.

123. W. Y. Yeong, C. K. Chua, K. F. Leong, M. Chandrasekaran: “Rapid

prototyping in tissue engineering: challenges and potential”, Trends in


124. B. S. Kim, D. J. Mooney: “Development of biocompatible synthetic

extracellular matrices for tissue engineering”, Trends in Biotechnology,

Vol. 16, No. 5, 1998, pp. 224-230.

125. A. Ovsianikov, M. Gruene, M. Pflaum, L. Koch, F. Maiorana, M.

Wilhelmi, A. Haverich, B. Chichkov: “Laser printing of cells into 3D

scaffolds”, Biofabrication, Vol. 2, No. 1, 2010, 014104.

Bibliography

171 of 176

126. H. Lo, M. S. Ponticiello, K. W. Leong: “Fabrication of controlled release

biodegradable foams by phase separation”, Tissue Engineering, Vol. 1.

No. 1, 1995, pp. 15-28.

127. A. G. Mikos, A. J. Thorsen, L. A. Czerwonka, Y. Bao, R. Langer, D. N.

Winslow, J. P. Vacanti: “Preparation and characterization of poly (L-

lactic acid) foams”, Polymer, Vol. 35, 1994, pp. 1068-1077.

128. D. J. Mooney, D. F. Baldwin, N. P. Suh, J. P. Vacanti, R. Langer: “Novel

approach to fabricate porous sponges of poly (D, L-lactic-co-glycolic acid)

without the use of organic solvents”, Biomaterials, Vol. 17, No. 14, 1996,

pp. 1417-1422.

129. K. Whang, C. H. Thomas, K. E. Healy, G. Nuber: “A novel method to

fabricate bioabsorbable scaffolds”, Polymer, Vol. 36, No. 4, 1995, pp.

837-842.

130. P. X. Ma, R. Y. Zhang: “Microtubular architecture of biodegradable

polymer scaffolds”, Journal of Biomedical Materials Research, Vol. 56,

No. 4, 2001, pp. 469-477.

131. D. W. Hutmacher: “Scaffold design and fabrication technologies for

engineering tissues state of the art and future perspectives”, Journal of

Biomaterials Science, Polymer Edition, Vol. 12, No. 1, 2001, pp. 107-124.

132. R. Landers, R. Mülhaupt: “Desktop manufacturing of complex objects,

prototypes and biomedical scaffolds by means of computer-assisted

design combined with computer-guided 3D plotting of polymers and

reactive oligomers”, Macromolecular Materials and Engineering, Vol. 282,

No. 1, 2000, pp. 17-21.

133. W. Sun, A. Darling, B. Starly, J. Nam: “Computer-aided tissue

engineering: overview, scope and challenges”, Biotechnology and

Applied Biochemistry, Vol. 39, 2004, pp. 29-47.

134. R. A. Barry, R. F. Shepherd, J. N. Hanson, R. G. Nuzzo, P. Wiltzius, J. A.

Lewis: “Direct-write assembly of 3D hydrogel scaffolds for guided cell

growth”, Advanced Materials, Vol. 21, No. 23, 2009, pp. 2407-2410.

135. T. Boland, X. Tao, B. J. Damon, B. Manley, P. Kesari, S. Jalota, S.

Bhaduri: “Drop-on-demand printing of cells and materials for designer

tissue constructs”, Materials Science and Engineering C: Biomimetic and

Supramolecular Systems, Vol. 27, No. 3, 2007, pp. 372-376.

136. T. Boland, V. Mironov, A. Gutowska, E. A. Roth, R. R. Markwald: “Cell

and organ printing 2: fusion of cell aggregates in three-dimensional gels”,

Anatomical Record Part A: Discoveries in Molecular Cellular and

Evolutionary Biology, Vol. 272A, No. 2, 2003, pp. 497-502.

Bibliography

172 of 176

137. V. Mironov, T. Boland, G. Forgacs, R. R. Markwald: “Organ printing:

computer-aided jet-based 3D tissue engineering”, Trends in


138. Website cited:

http://images.google.com.sg/imglanding?q=organ%20printing&imgurl=ht

tp://www.newscientist.com/data/images/ns/cms/dn3292/dn3292-

2_587.jpg&imgrefurl=http://3d-

print.in/%3Fp%3D10&usg=__x2XuY73RdyXWWQpViuXNmmpmDrI=

&h=388&w=587&sz=35&hl=en&um=1&itbs=1&tbnid=0j4Vymt6KgiHh

M:&tbnh=89&tbnw=135&prev=/images%3Fq%3Dorgan%2Bprinting%2

6um%3D1%26hl%3Den%26sa%3DN%26rlz%3D1R2ADFA_enSG346%

26tbs%3Disch:1&um=1&sa=N&rlz=1R2ADFA_enSG346&tbs=isch:1&s

tart=11#tbnid=0j4Vymt6KgiHhM&start=15

139. S. Moon, S. K. Hasan, Y. S. Song, F. Xu, H. O. Keles, F. Manzur, S.

Mikkilineni, J. W. Hong, J. Nagatomi, E. Haeggstrom, A.

Khademhosseini, U. Demirci: “Layer by layer three-dimensional tissue

epitaxy by cell-laden hydrogel droplets”, Tissue Engineering Part C:

Methods, Vol. 16, No. 1, 2010, pp. 157-166.

140. E.Q. Li, S.T. Thoroddsen, J.Y.H. Fuh, S.C.H. Thian, Y.S. Wong, H.T.

Loh, L. Lu, PET-based piezoelectric squeeze mode microjetting printhead

with interchangeable nozzles, Provisional US Patent (2009) Serial No.

61/226781.

141. Y. Nishiyama, M. Nakamura, C. Henmi, K. Yamaguchi, S. Mochizuki, H.

Nakagawa, K. Takiura: “Development of a three-dimensional bioprinter:

construction of cell supporting structures using hydrogel and state-of-the-

art inkjet technology”, Journal of Biomechanical Engineering -

Transaction of The ASME, Vol. 131, No. 3, 2009, 035001.

142. J. D. Kim, J. S. Choi, B. S. Kim, Y. C. Choi, Y. W. Cho: “Piezoelectric

inkjet printing of polymers: stem cell patterning on polymer substrates”,

Polymer, 2010, article in press.

143. S. Ilkhanizadeh, A. I. Teixeira, O. Hermanson: “Inkjet printing of

macromolecules on hydrogels to steer neural stem cell differentiation”,

Biomaterials, Vol. 28, No. 27, 2007, pp. 3936-3943.

144. E. A. Roth, T. Xu, M. Das, C. Gregory, J. J. Hickman, T. Boland: “Inkjet

printing for high-throughput cell patterning”, Biomaterials, Vol. 25, No.

17, 2004, pp. 3707-3715.

145. N. E. Sanjana, S. B. Fuller: “A fast flexible ink-jet printing method for

patterning dissociated neurons in culture”, Journal of Neuroscience

Methods, Vol. 136, No. 2, 2004, pp. 151-163.

146. G. M. Whitesides, B. Grzybowski: “Self-assembly at all scales”, Science,

Vol. 295, No. 5564, 2002, pp. 2418-2421.

http://images.google.com.sg/imglanding?q=organ%20printing&imgurl=http://www.newscientist.com/data/images/ns/cms/dn3292/dn3292-2_587.jpg&imgrefurl=http://3d-print.in/%3Fp%3D10&usg=__x2XuY73RdyXWWQpViuXNmmpmDrI=&h=388&w=587&sz=35&hl=en&um=1&itbs=1&tbnid=0j4Vymt6KgiHhM:&tbnh=89&tbnw=135&prev=/images%3Fq%3Dorgan%2Bprinting%26um%3D1%26hl%3Den%26sa%3DN%26rlz%3D1R2ADFA_enSG346%26tbs%3Disch:1&um=1&sa=N&rlz=1R2ADFA_enSG346&tbs=isch:1&start=11#tbnid=0j4Vymt6KgiHhM&start=15









Bibliography

173 of 176

147. Z. C. Liu, Y. Su, K. Varahramyan: “Inkjet-printed silver conductors using

silver nitrate ink and their electrical contacts with conducting polymers”,

Thin Solid Films, Vol. 478, No. 1-2, 2005, pp. 275–279.

148. W. Feng, J. Y. H. Fuh, Y. S. Wong: “Development of a drop-on-demand

micro dispensing system”, Materials Science Forum, Vol. 505-507, 2006,

pp. 25-29.

149. B. M. Wu, S. W. Borland, R. A. Giordano, L. G. Cima, E. M. Sachs, M. J.

Cima: “Solid free-form fabrication of drug delivery devices”, Journal of

Controlled Release, Vol. 40, No. 1-2, 1996, pp. 77-87.

150. T. Kawase, T. Shimoda, C. Newsome, H. Sirringhaus, R. H. Friend:

“Inkjet printing of polymer thin film transistors”, Thin Solid Films, Vol.

438, 2003, pp. 279-287.

151. K. X. Steirer, J. J. Berry, M. O. Reese, M. F. A. M. van Hest, A. Miedaner,

M. W. Liberatore, R. T. Collins, D. S. Ginley: “Ultrasonically sprayed

and inkjet printed thin film electrodes for organic solar cells”, Thin Solid

Films, Vol. 517, No. 8, 2009, pp. 2781-2786.

152. D. Xu, V. Sanchez-Romaguera, S. Barbosa, W. Travis, J. de Wit, P. Swan,

S. G. Yeates: “Inkjet printing of polymer solutions and the role of chain

entanglement”, Journal of Materials Chemistry, Vol. 17, No. 46, 2007, pp.

4902-4907.

153. K. C. Fan, J. Y. Chen, C. H. Wang, W. C. Pan: “Development of a drop-

on-demand droplet generator for one-drop-fill technology”, Sensors and

Actuators: A - Physical, Vol. 147, No. 2, 2008, pp. 649-655.

154. J. de Jong, G. de Bruin, H. Reinten, M. van den Berg, H. Wijshoff, M.

Versluis, D. Lohse: “Air entrapment in piezo-driven inkjet printheads”,

Journal of Acoustical Society of America, Vol. 120, No. 3, 2006, pp.

1257-1265.

155. J. de Jong, H. Reinten, H. Wijshoff, M. van den Berg, K. Delescen, R. van

Dongen, F. Mugele, M. Versluis, D. Lohse: “Marangoni flow on an inkjet

nozzle plate”, Applied Physics Letters, Vol. 91, No. 20, 2007, 204102.

156. B. Beulen, J. de Jong, H. Reinten, M. van den Berg, H. Wijshoff, R. van

Dongen: “Flows on the nozzle plate of an inkjet printhead”, Experiments

in Fluids, Vol. 42, No. 2, 2007, pp. 217-224.

157. B. H. Kim, T. G. Kim, T. K. Lee, S. Kim, S. J. Shin, S. J. Kim, S. J. Lee:

“Effects of trapped air bubbles on frequency responses of the piezo-driven

inkjet printheads and visualization of the bubbles using synchrotron X-

ray”, Sensors and Actuators A: Physical, Vol. 154, No. 1, 2009, pp. 132-

139.

Bibliography

174 of 176

158. S. J. Lee, D. H. Kwon, Y. S. Choi: “Dynamics of entrained air bubbles

inside a piezodriven inkjet printhead”, Applied Physics Letters, Vol. 95,

No. 22, 2009, 221902.

159. J. de Jong, R. Jeurissen, H. Borel, M. van den Berg, H. Wijshoff, H.

Reinten, M. Versluis, A. Prosperetti, D. Lohse: “Entrapped air bubbles in

piezo-driven inkjet printing: Their effect on the droplet velocity”, Physics

of Fluids, Vol. 18, No. 12, 2006, 121511.

160. H. Wijshoff: “Structure- and fluid-dynamics in piezo inkjet printheads”,

Ph.D. Thesis, University of Twente, The Netherlands, 2008.

161. K.S. Kwon: “Methods for detecting air bubble in piezo inkjet dispensers”,

Sensors and Actuators A: Physical, Vol. 153, 2009, pp. 50-56.

162. A. M. Worthington: “A study of Splashes”, Longmans and Green, London,

1908; reprint by Macmillan, Basingstoke, 1963.

163. M. Rein: “Phenomena of liquid drop impact on solid and liquid surfaces,”

Fluid Dynamics Research, Vol. 12, No. 2, 1993, pp. 61-93.

164. S. T. Thoroddsen, A. Q. Shen: “Granular jets,” Physics of Fluids, Vol. 13,

No. 1, 2001, pp. 4-6.

165. D. Lohse, R. Bergmann, R. Mikkelsen, C. Zeilstra, D. van der Meer, M.

Versluis, K. van der Weele, M. van der Hoef, H. Kuipers: “Impact on soft

sand: Void collapse and jet formation,” Physical Review Letters, Vol. 93,

No. 19, 2004, 198003.

166. J. R. Royer, E. I. Corwin, A. Flior, M.-L. Cordera, M. L. Rivers, P. J. Eng,

H. M. Jeager: “Formation of granular jets observed by high-speed x-ray

radiography,” Nature Physics, Vol. 1, 2005, pp. 164-167.

167. A. Antkowiak, N. Bremond, J. Duplat, S. Le Dizes, E. Villermaux:

“Cavity jets”, Physics of Fluids, Vol. 19, No. 9, 2007, 091112.

168. A. Antkowiak, N. Bremond, S. Le Dizes, E. Villermaux: “Short-term

dynamics of a density interface following an impact”, Journal of Fluid

Mechanics, Vol. 577, 2007, pp. 241-250.

169. J. O. Marston, S. T. Thoroddsen: “Apex jets from impacting drops”, The

Journal of Fluid Mechanics, Vol. 614, 2008, pp. 293-302.

170. S. T. Thoroddsen, K. Takehara, T. G. Etoh, C. D. Ohl: “Spray and

microjets produced by focusing a laser into a hemispherical drop”,

Physics of Fluids, Vol. 21, No. 11, 2009, 112101.

171. S. T. Thoroddsen, T. G. Etoh, K. Takehara: “Microjetting from wave

focusing on oscillating drops”, Physics of Fluids, Vol. 19, No. 5, 2007,

052101.

Bibliography

175 of 176

172. A. L. Yarin: “Drop impact dynamics: Splashing, spreading, receding,

bouncing”, Annual Review of Fluid Mechanics, Vol. 38, 2006, pp. 159-

192.

173. D. Bartolo, C. Josserand, D. Bonn: “Singular jets and bubbles in drop

impact”, Physical Review Letters, Vol. 96, No. 12, 2006, 124501.

174. B. W. Zeff, B. Kleber, J. Fineberg, D. P. Lathrop: “Singularity dynamics

in curvature collapse and jet eruption on a fluid surface”, Nature, Vol. 403,

No. 6768, 2000, pp. 401-404.

175. M. P. Brenner: “Fluid mechanics - jets from a singular surface”, Nature,

Vol. 403, No. 6768, 2000, pp. 377-378.

176. T. G. Etoh, D. Poggemann, G. Kreider, H. Mutoh, A. J. P. Theuwissen, A.

Ruckelshausen, Y. Kondo, H. Maruno, K. Takubo, H. Soya, K. Takehara,

T. Okinaka, Y. Takano: “An image sensor which captures 100

consecutive frames at 1,000,000 frames/s”, IEEE Transactions on

Electron Devices, Vol. 50, No. 1, 2003, pp. 144-151.

177. M. S. Barrow, S. W. J. Brown, P. R. Williams: “Extensional flow of

liquid jets formed by bubble collapse in oils under cavitation-generated

pressure waves”, Experiments in Fluids, Vol. 36, 2004, pp. 463-472.

178. S. T. Thoroddsen, T. G. Etoh, K. Takehara, N. Ootsuka: “On the

coalescence speed of bubbles”, Physics of Fluids, Vol. 17, 2005, 071703.

179. F. H. Zhang, S. T. Thoroddsen: “Satellite generation during bubble

coalescence”, Physics of Fluids, Vol. 20, No. 2, 2008, 022104.

180. L. Duchemin, S. Popinet, C. Josserand, S. Zaleski: “Jet formation in

bubbles bursting at a free surface”, Physics of Fluids, Vol. 14, No. 9, 2002,

pp. 3000-3008.

181. G. K. Batchelor: “An introduction to fluid dynamics”, London:

Cambridge University Press, 1967.

182. A. Papoulis: “Probability, random variables, and stochastic processes”, 1st

Edition, New York: McGraw-Hill, 1965.

183. C. K. Kuo, P. X. Ma: “Ionically corsslinked alginate hydrogels as

scaffolds for tissue engineering: Part 1. Structure, gelation rate and

mechanical properties”, Biomaterials, Vol. 22, 2001, pp. 511-521.

Publications

176 of 176

Publications

1. E. Q. Li, S. T. Thoroddsen, J. Y. H. Fuh, S. C. H. Thian, Y. S. Wong, H. T.

Loh, L. Lu, PET-based piezoelectric squeeze mode microjetting printhead

with interchangeable nozzles, Provisional US Patent (2009) Serial No.

61/226781.

2. F. H. Zhang, E. Q. Li, S. T. Thoroddsen: “Satellite formation during

coalescence of unequal size drops”, Physics Review Letters, Vol. 102,

104502, 2009.

3. E. Q. Li, J. Y. H. Fuh, Y. S. Wong, S. T. Thoroddsen: “Forming a fine jet in

inkjet printing”, American Physical Society, 62nd

Annual Meeting of the

APS Division of Fluid Dynamics, November 22-24, 2009.

4. E. Q. Li, Q. Xu, J. Sun, J. Y. H. Fuh, Y. S. Wong, S. T. Thoroddsen:

“Design and fabrication of a PET/PTFE-based piezoelectric squeeze mode

drop-on-demand inkjet printhead with interchangeable nozzle”, Sensors

and Actuators A: Physical, article accepted, 2010.

5. E. Q. Li, E. K. Tan, S. T. Thoroddsen: “Piezoelectric Drop-on-Demand

Inkjet Printing of Rat Fibroblast Cells: Survivability Study and Pattern

Printing”, Biotechnology and Bioengineering, 2011. (In Preparation)

THE GENERATION AND EXPERIMENTAL STUDY OF MICROSCALE ...

Documents